INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-501 Course Title: Essential Mathematics for AI

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Autumn 7. Subject Area: PCC
8. Pre-requisite: Nil
9. Objective: To introduce students to the various Mathematical concepts to be used in ML and AI

10. Details of the Course

S.No. Contents Contact

hours
1. Basics of Linear Algebra: System of Linear Equations; Vector space and sub- 10
spaces (definition, examples and concepts of basis); Linear mappings; Matrices;
Eigenvalues and Eigenvectors Norms; Inner Product; Orthogonally; Spectral
Decomposition; Singular value Decomposition; Low-rank Approximation;
Projection; Principal Component Analysis and Generative Models
2. Gradient Calculus: Differentiation of univariate functions; partial derivatives and 6
gradients; gradients of vector valued functions and matrices; Backpropagation and
automatic differentiation; Linearization and Multivariate Taylor Series
3. Optimization: Notion of maxima and minima; Optimization using gradient 8
descent; Constrained Optimization techniques; Convex optimization Algorithms
4. Probability and Statistics: Basic concepts of probability: conditional probability, 14
Bayes’ theorem, random variables, moments, moment generating functions, some
useful distributions, Joint distribution, conditional distribution, transformations of
random variables, covariance, correlation, random sample, statistics, sampling
distributions, point estimation, MAP, MLE
5. Information Theory: Entropy, cross-entropy, KL divergence, mutual information; 4
Total 42
11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/Reprint
1 M. P. Deisenroth, A. A. Faisal, C. S. Ong, Mathematics for 2020
Machine Learning, Cambridge University Press (1st edition)
2 S. Axler, Linear Algebra Done Right. Springer International 2015
Publishing (3rd edition)
3 J. Nocedal and S. J. Wright, Numerical Optimization. New York: 2006
Springer Science+Business Media
4 E. Kreyszig, Advanced Engineering Mathematics, John Wiley 2015
and Sons, Inc., U.K. (10th Edition)
5 R. A. Johnson, I. Miller, and J. E.Freund, "Miller & Freund’s 2011
Probability and Statistics for Engineers", Prentice Hall PTR, (8th
edition)
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-503 Course Title: Computer Architecture for AI

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Autumn 7. Subject Area: PCC
8. Pre-requisite: Nil
9. Objective: To learn the design of hardware architectures and accelerators for deep-learning/artificial-
intelligence. This course is at the intersection of deep-learning and computer-architecture/embedded-
system/VLSI.

10. Details of the Course:

S.No. Contents Contact

Hours
1. Background topics: Approximate computing and storage, Roofline Model, 8
Cache tiling (blocking), GPU architecture, CUDA programming, understanding
GPU memory hierarchy, FPGA architecture, Matrix multiplication using systolic
array
2. Convolutional strategies: Direct, FFT-based, Winograd-based and Matrix- 3
multiplication based.
3. Deep learning on various hardware platforms: Deep learning on FPGAs and 15
case study of Microsoft's Brainwave, Deep learning on Embedded System
(especially NVIDIA's Jetson Platform), Deep learning on Edge Devices
(smartphones), Deep learning on an ASIC (especially Google's Tensor Processing
Unit.), Deep-learning on CPUs and manycore processor (e.g., Xeon Phi),
Memristor-based processing-in-memory accelerators for deep-learning.
4 Memory-efficiency and reliability of DNN accelerators: Model-size aware 6
Pruning of DNNs, Hardware architecture-aware pruning of DNNs, Understanding
soft-errors. Understanding reliability of deep learning algorithms and accelerators
5 Memory-related tradeoffs in DNN accelerators: Comparison of memory 4
technologies (SRAM, DRAM, eDRAM, STT-RAM, PCM, Flash) and their
suitability for designing memory-elements in DNN accelerator, Neural branch
predictors and their applications
6 Autonomous driving and DNN training: Hardware/system-challenges in 6
autonomous driving, Distributed training of DNNs and addressing memory
challenges in DNN training
Total 42
11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/Reprint
1. Hennessy, J. L. ,& Patterson, D. A., Computer Architecture: A 2017
quantitative approach (Sixth Edition), Elsevier
https://www.google.co.in/books/edition/Computer_Architecture/cM8m
DwAAQBAJ
2. Brandon Reagen, Robert Adolf, Paul Whatmough, Gu-Yeon Wei, and 2017
David Brooks Deep Learning for Computer Architects Synthesis
Lectures on Computer Architecture, August 2017, Vol. 12, No. 4,
Pages 1-123
(https://doi.org/10.2200/S00783ED1V01Y201706CAC041)
3 Tor M. Aamodt, Wilson Wai Lun Fung, and Timothy G. Rogers 2018
General- Purpose Graphics Processor Architectures, Synthesis
Lectures on Computer Architecture, May 2018, Vol. 13, No. 2 , Pages
1-140 (https://doi.org/10.2200/S00848ED1V01Y201804CAC044)
4 Goodfellow, I., Bengio, Y., Courville, A., & Bengio, Y. (2016). 2016
Deeplearning (Vol. 1, No. 2). Cambridge: MIT press.
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Center for Artificial Intelligence and Data Science

1. Subject Code: AID-505 Course Title: Machine Learning

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Autumn 7. Subject Area: PCC
8. Pre-requisite: Nil
9. Objective: To provide an understanding of the theoretical concepts of machine learning and prepare
students for research or industry application of machine learning techniques.

10. Details of the Course

S.No. Contents Contact

hours
1. Introduction: Well-posed learning problems, examples of machine learning
applications, model selection and generalization, concept learning, inductive 4
learning hypothesis, inductive bias. Information theory: entropy, mutual
information, KL divergence
2. Performance Optimization: Directional Derivatives, Minima, Necessary
Conditions for Optimality, Convex function, Gradient Descent, Stable learning 4
rates, Newtons Method, Conjugate gradient method, The Levenberg-Marquardt
algorithm.
3. Linear Classification: Linear classifier, Logistic Regression, Decision Boundary,
Cost Function Optimization, Multi-class Classification, Bias and Variance, L1 and 4
L2 Regularization, feature reduction, Principal Component Analysis, Singular
Value Decomposition
4. Artificial Neural Networks: Perceptron, Linear Networks, Multi-layer Networks,
Forward propagation, Backward propagation, Alternative activation functions, 5
variations on backpropagation, Deep neural networks.
5. Decision tree learning: Decision tree representation, appropriate problems for
decision tree learning, hypothesis space search in decision tree learning, inductive 5
bias in tree learning, avoiding overfitting the data, alternative measures for
selecting attribute values, ensemble methods, bagging, boosting, random forest
6. Support Vector Machines: Computational learning theory, probably
approximately correct (PAC) learning, sample complexity and VC dimension, 5
linear SVM, soft margin SVM, kernel functions, nonlinear SVM, Multiclass
classification using SVM, Support vector regression.
Instance based learning: K-nearest neighbor learning, distance weighted neighbor
7. learning, locally weighted regression, adaptive nearest neighbor methods, The 3
Concept of Unsupervised Learning, Competition networks, K-means clustering
algorithm.
 8. Bayesian Learning: Bayes theorem, maximum likelihood and least squared error
hypotheses, Naive Bayes classifier, Bayesian belief networks, gradient ascent 7
training of Bayesian networks, learning the structure of Bayesian networks, the EM
algorithm, mixture of models, Markov models, hidden Markov models.
9. Reinforcement learning: the learning task, Q learning, convergence, temporal
difference learning, nondeterministic rewards and actions, generalization, 5
relationship to dynamic programming.
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication /Reprint

1. T. Mitchell, Machine Learning, McGraw Hill 1997

2. Christopher Bishop, Pattern Recognition and Machine Learning, 2006
Springer
3. K. Murphy. Machine Learning: A probabilistic perspective, MIT 2012
Press
4. Hastie, Tibshirani, Friedman, Elements of statistical learning, 2011
Springer
5. I. Goodfellow, Y. Bengio and A. Courville. Deep Learning. MIT 2016
Press
6. Richard S. Sutton and Andrew G. Barto, Reinforcement Learning: 2018
AnIntroduction, MIT Press
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-507 Course Title: Advanced Data Structures and Algorithms
2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Autumn 7. Subject Area: PCC
8. Pre-requisite: Nil
9. Objective: To introduce advanced concepts in data structures and algorithms.

10. Details of the Course

S.No. Contents Contact

hours
1. Data Structures: Priority queues and heaps, dictionaries, hash tables, binary search 8
trees, interval trees
2. Basic Algorithms: Asymptotic notation, recursion, divide-and-conquer paradigm, 8
greedy strategy, dynamic programming, graph algorithms, complexity classes P,
NP, NP-hard, NP-complete.
3. Approximation Algorithms: Performance ratio, vertex cover problem, travelling 8
salesman problem, set covering problem, subset sum problem.
4. Randomized Algorithms: Tools and techniques. Applications. 8
5. Multithreaded Algorithms: Dynamic multithreaded programming, multithreaded 10
matrix multiplication, multithreaded merge sort.
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/ Reprint
1. Wirth, N., “Algorithms and Data Structures”, Prentice-Hall of India. 2017
2. Motwani and Raghavan, Randomized Algorithms. Cambridge 2014
University Press.
3. Brad Miller and David Ranum, Luther College, “Problem 2013
Solving with Algorithms and Data Structures Using Python,” Franklin,
Beedle & Associates
4. Cormen T, Introduction to Algorithms, MIT Press, 3rd Edition. 2009
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-509 Course Title: Programming for AI

2. Contact Hours: L: 0 T: 0 P: 4
3. Examination Duration (Hrs.): Theory: 0 Practical: 2
4. Relative Weightage: CWS: 0 PRS: 50 MTE: 0 ETE: 0 PRE: 50
5. Credits: 2 6. Semester: Autumn 7. Subject Area: PCC
8. Pre-requisite: Nil
9. Objective: This course's objective is to provide hands-on experience on the various programming
components for Artificial Intelligence.

10. Details of the Course:

S.No. Contents Contact
hours
1. Python: Basics, Numpy, Pandas, and Matplotlib 16
2. Scikit-Learn and NLTK 12
3. TensorFlow and Keras 12
TensorFlow Lite: Deploy machine learning systems on IoT device (Arduino
4. Platform and Raspberry Pi based devices) (C/C++, Python) 16
Total 56

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of
Publication/ Reprint
Jake VanderPlas “Python Data Science Handbook,” First
1 2016
Edition, O'Reilly Media,Inc.
2 Wes McKinney “Python for Data Analysis: Data Wrangling 2017
with Pandas, NumPy, and IPython,” Second Edition, O'Reilly
Media, Inc.
3 Pramod Singh and Avinash Manure “Learn TensorFlow 2.0: 2020
Implement Machine Learning and Deep Learning Models with
Python,” First Edition, Apress
4 Aurélien Géron “Hands-On Machine Learning with Scikit- 2019
Learn, Keras, and TensorFlow,” Second Edition, O'Reilly
Media, Inc.
5 J. M. Hughes “Arduino: A Technical Reference: A Handbook 2016
for Technicians, Engineers, and Maker,” First Edition, O'Reilly
Media, Inc.
6 Derek Molloy “Exploring Raspberry Pi: Interfacing to the Real 2016
World with Embedded Linux,” First Edition, Wiley
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-551 Course Title: Convex Optimization in Machine Learning

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: NIL
9. Objective: To introduce convex optimization algorithms to be used in various machine learning tools.

10. Details of the Course

S. No. Contents Contact

hours
1. Introduction: significance of optimization methods in machine learning, a brief 8
review of the fundamentals of optimization, Convex sets and convex functions,
Problems in Convex Optimization (linear/quadratic/Semi-definite programming),
Strong and weak duality, rates of convergence
2. Optimization models: Types of optimization models arising in different areas of 6
ML, large scale optimization
3. First order optimization methods: Gradient descent, stochastic gradient descent, 8
NAG, Adam, ADMM, Frank and Wolfe, SVRG, AdaGrad, Implementation of
these algorithms and their advantages and disadvantages
4. Second and higher order optimization methods: Conjugate gradient, Newton’s 8
method, Quasi newton method, stochastic quasi Newton method, Hessian free
method, Natural Gradient Method, Implementation of these algorithms and their
advantages and disadvantages.
5. Optimization Solvers and Toolboxes: CVX (MATLAB), CVXPY (Python), 6
CVXOPT (Python)
6. Case Studies: Recent developments and advanced optimization algorithms 6
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication / Reprint
1. Stephen Boyd and Lieven Vandenberghe, “Convex Optimization”. 2004
Cambridge University Press,
2. Suvrit Sra, Sebestian Nowozin and Stephen J. Wright, “Optimization 2013
for Machine Learning”, PHI
3. Neal Parikh and Stephen Boyd, Proximal Algorithms, NOW 2013
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Department of Computer Science and Engineering

1. Subject Code: CSN-515 Course Title: Data Mining and Warehousing

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Spring 7. Subject Area: PEC
8. Pre-requisite: CS-102
9. Objective: To educate students to the various concepts, algorithms and techniques in data mining and
warehousing and their applications.

10. Details of the Course

S.No. Contents Contact

hours
1. Introduction to data mining: Motivation and significance of data mining, data 3
mining functionalities, interestingness measures, classification of data mining
system, major issues in data mining.
2. Data pre-processing: Need, data summarization, data cleaning, data integration 6
and transformation, data reduction techniques —Singular Value Decomposition
(SVD), Discrete Fourier Transform (DFT), Discrete Wavelet Transform
(DWT), data discretization and concept hierarchy generalization.
3. Data warehouse and OLAP technology: Data warehouse definition, 4
multidimensional data model(s), data warehouse architecture, OLAP server
types,data warehouse implementation, on-line analytical processing and mining,
4. Data cube computation and data generalization: Efficient methods for data 4
cube computation, discovery driven exploration of data cubes, complex
aggregation, attribute oriented induction for data generalization.
5. Mining frequent patterns, associations and correlations: Basic concepts, 6
efficient and scalable frequent itemset mining algorithms, mining various kinds
of association rules —multilevel and multidimensional, association rule mining
versus correlation analysis, constraint based association mining.
6. Classification and prediction: Definition, decision tree induction, Bayesian 6
classification, rule based classification, classification by backpropagation and
support vector machines, associative classification, lazy learners, prediction,
accuracy and error measures.
7. Cluster Analysis: Definition, Clustering Algorithms - partitioning, hierarchical, 6
density based, grid based and model based; Clustering high dimensional data,
constraint based cluster analysis, outlier analysis - density based and distance
based.
8. Data mining on complex data and applications: Algorithms for mining 7
of spatial data, multimedia data, text data: data mining applications, social
impactsof data mining, trends in data mining.
Total 42
11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication / Reprint
1. Marakas, George M. Modern data warehousing, mining, and 2003
visualization: core concepts. Upper Saddle River, NJ: Prentice
Hall, 2003.
2. Pujari, Arun K. Data mining techniques. Universities press, 2001. 2001
3. Lee, Mong Li, Hongjun Lu, Tok Wang Ling, and Yee Teng Ko. 1999
"Cleansing data for mining and warehousing." In International
Conference on Database and Expert Systems Applications, pp.
751-760. Springer, Berlin, Heidelberg, 1999.
4. Wang, John, ed. Encyclopedia of data warehousing and mining. iGi 2005
Global, 2005.
5 Gupta, Gopal K. Introduction to data mining with case studies. PHI 2014
Learning Pvt. Ltd., 2014.
6 Tan, Pang-Ning, Michael Steinbach, and Vipin Kumar. 2016
Introduction to data mining. Pearson Education India, 2016.
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-552 Course Title: Deep Learning

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Machine Learning
9. Objective: The objective of this course is to learn deep learning algorithms, concepts, experiments,
research along with their application on generic use cases.

10. Details of the Course:

S.No. Contents Contact
Hours
1 Introduction to deep learning, logical computations with neurons, perceptron, 6
backpropagation, historical trends, applications, and use-cases for industry
2 Deep Networks: Training a deep neural network (DNN), hidden layers, activation 7
functions, fine-tuning neural network hyper-parameters
3 Custom Deep Neural Networks: vanishing/exploding gradient issues, reusing pre- 8
trained layers, optimizers, l1 and l2 regularization, dropout
4 Convolutional neural networks (CNNs): convolutional layer, filters, stacking, 7
pooling layer, CNN architectures
5 Recurrent neural networks (RNNs): recurrent neurons, unrolling, input and output 7
sequences, training RNNs, deep RNNs, LSTM cell, GRU cell
6 Representation Learning and Generative Learning: Auto encoders: data 7
representations, linear auto encoder, stacked auto encoders, variational auto
encoders
Total 42
11. Suggested Books:
S.No. Name of Authors/Book/Publisher Year of
Publication/Reprint
Aurélien Géron, “Hands-On Machine Learning with Scikit-Learn,
1 Keras, and TensorFlow: Concepts, Tools, and Techniques to Build 2019
Intelligent Systems,”Second Edition, O'Reilly Media
2 Ian Goodfellow, Yoshua Bengio, and Aaron Courville, “Deep 2017
Learning,” FirstEdition, MIT Press
3 François Chollet “Deep Learning with Python,” First Edition, Manning 2018
Publication
4 Rowel Atienza “Advanced Deep Learning with Keras,” First Edition, 2018
Packt Publishing
5 Sudharsan Ravichandran “Hands-On Deep Learning Algorithms with 2019
Python,” First Edition, Packt Publishing
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-553 Course Title: Digital Image Processing

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weight: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: The objective of this course is to introduce the fundamental techniques and
algorithms used for acquiring, processing and extracting useful information from digital images.

10. Details of the Course

S.No. Contents Contact

Hours
1. Introduction: Signal processing overview; Image processing basics; 4
Fundamental signals (1-D and 2-D); Classification of systems; Characteristics
of LTI/LSI systems. Introduction to the DIP areas and applications.
2. Digital Image Fundamentals: Human visual system and visual 4
perception; Image sensing and acquisition Image file types; Pixel
representation and spatial relationship
3. Image Digitization: Sampling and quantization. Image Transforms: 2- D 8
DSFT and 2-D DFT, 2-D discrete cosine transform (DCT), 1-D and 2-D
Karhonen Loeve (KL) or principal component analysis (PCA) and 1-D and 2-
D discrete wavelet transforms and relation to filter banks.
4. Image Enhancement: Point and algebraic operations, edge detection and 6
sharpening, filtering in the spatial and transformed domains. Rotation,
interpolation, image filtering, spatial operators, morphological operators.
5. Image Segmentation: Thresholding; Edge based segmentation; Region 6
growing; Watershed transform.
Image Restoration: Degradation models, inverse and pseudo-inverse
filtering, 2-D Wiener filtering and implementation
6. Image Compression and Encoding: Entropy-based schemes, Transform- based 4
encoding, Predictive encoding and DPCM, Vector quantization, Huffman
coding.
7. Feature Extraction and Segmentation: Contour and shape dependent feature 5
extraction, textural features, region-based and feature-based segmentation.
8. Pattern Classification: Standard linear and Bayesian classifiers, 5
supervised Vs unsupervised classification, classification performance index.
Applications in satellite, sonar, radar and medical areas.
Total 42
11. Suggested Books:

S.No. Name of Authors /Books / Publisher Year of

Publication/Reprint
1. Gonzalez R. C. and Woods R. E., “Digital image processing,” 2017
FourthEdition, Prentice Hall.
2. Lim J. S., “Two-dimensional signal and image processing,” Prentice 1990
Hall.
3. Dudgeon D.E. and Merserau R. M., “Multidimensional digital 1984
signal processing,” Prentice Hall Signal Processing Series.
4. Bose T., “Digital Signal and Image Processing”, Wiley India. 2010
5. Sonaka M., Hlavac V. and Boyle R., “Image Processing, 2017
Analysis and Machine Vision,” Fourth edition, Cengage India
Private Limited.
6. W. K. Pratt. “Digital Image Processing,” Fourth Edition, John 2007
Wiley &Sons, New York.
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Department of Mathematics

1. Subject Code: MAN-628 Course Title: Evolutionary Algorithms

2. Contact Hours: L: 3 T: 0 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 3 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: To provide knowledge about basic concepts of Evolutionary Algorithms
10. Details of the Course:

S.No. Contents Contact

Hours
1 Genetic Algorithms: Historical development, GA concepts – encoding, 12
fitness function, population size, selection, crossover and mutation operators,
along with the methodologies of applying these operators. Binary GA and their
operators, Real Coded GA and their operators
2 Particle Swarm Optimization: PSO Model, global best, Local best, velocity 10
update equations, position update equations, velocity clamping, inertia weight,
constriction coefficients, synchronous and asynchronous updates, Binary
PSO.
3 Memetic Algorithms: Concepts of memes, Incorporating local search as 5
memes, single and multi-memes, hybridization with GA and PSO, Generation
Gaps, Performance metrics.
4 Differential Evolution: DE as modified GA, generation of population, 5
operators and their implementation.
5 Artificial Bee Colony: Historical development, types of bees and their role 5
in the optimization process.
6 Multi-Objective Optimization: Linear and nonlinear multi-objective 5
problems, convex and non – convex problems, dominance – concepts and
properties, Pareto – optimality, Use of Evolutionary Computations to solve
multi objective optimization, bi level optimization, Theoretical Foundations
Total 42

11. Suggested Books:

S.No. Name of Authors /Books / Publisher Year of

Publication/Reprint
Coello, C. A., Van Veldhuizen, D.A. and Lamont, G.B.: 2002
1. “Evolutionary Algorithms for solving Multi Objective Problems”,
Kluwer.
2. Deb, K.: “Multi-Objective Optimization using Evolutionary 2002
Algorithms”, John Wiley and Sons.
3. Deb, K.: “Optimization for Engineering Design Algorithms and 1998
Examples”, Prentice Hall of India.
4 Gen, M. and Cheng, R.: “Genetic Algorithms and Engineering 1997
Design”, Wiley, New York.
 Hart, W.E., Krasnogor, N. and Smith, J.E. : “Recent Advances in 2005
5. Memetic Algorithms”, Springer Berlin Heidelberg, New York.
6. Michalewicz, Z.: “Genetic Algorithms+Data tructures=Evolution 1992
Programs”, Springer-Verlag, 3rd edition, London, UK.
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Department of Computer Science and Engineering

1. Subject Code: CSN-528 Course Title: Natural Language Processing

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Basic knowledge of Artificial Intelligence
9. Objective: To provide an understanding of the theoretical concepts of Natural Language Processing
and prepare students for research or industry application of Natural Language Processing.

10. Details of the Course

S.No. Contents Contact

hours
1. Introduction to NLP, Corpus, Representation of Words, Preprocessing, Linguistic 6
and Statistical Properties of Words, POS Tagging, Parsing, Performance
Measures, Error Analysis, Confusion Matrix
2. Probability and NLP, n-Gram, Language Model, Join and Conditional 6
Probability, Chain Rule, Markov Assumption, Data Sparsity, Smoothing
Techniques, Generative Models, Naive Bayes
3. Distributed representation of words for NLP, Co-occurrence Matrix, 6
Collocations, Dimensionality Reduction, Singular Value Decomposition
4. Document Similarity, Inverted Index, Word2Vec, C-BoW, Skip-Gram Model, 6
Sampling, Hierarchical Soft-max, Sequence Learning
5. Neural Networks for NLP, Multi-Layer Perceptron, Activation Function, 6
Gradient Descent, Sequence Modeling, Recurrent Neural Networks
6. Gated Recurrent Unit, Long-Short Term Memory Networks, 1-D Convolutional 6
Layer, Language Model using RNN, Forward Pass, Backward Pass
7. Applications of NLP, Topic Modeling, Sentiment Analysis, Query Processing, 6
ChatBoat, Machine Translation, Statistical Machine Translation, Neural Machine
Translation, Spell Checker, Summarization
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication / Reprint
1. Manning, Christopher, and Hinrich Schutze. Foundations of 1999
statistical natural language processing. MIT press
2. Jurafsky, Dan. Speech & language processing. Pearson Education 2000
India
3. Smith, Noah A. Linguistic structure prediction. Morgan and 2011
Claypool
4. Kennedy, Graeme. An introduction to corpus linguistics. 2014
Routledge
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Department of Mathematics

1. Subject Code: MAN-653 Course Title: Numerical Optimization

2. Contact Hours: L: 3 T: 0 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 3 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: To provide knowledge about basic concepts of Numerical Optimization.
10. Details of the Course:
S.No. Contents Contact
Hours
1. Linear Programming: Review of various methods of linear programming 5
2. Nonlinear Programming 1-D Unconstrained Minimization Methods: Golden 6
Section, Fibonnacci Search, Bisection, Newton's Methods.
3. Multi-dimensional Unconstrained Minimization Methods: Cyclic Coordinate 10
Method, Hookes & Jeeves continuous and discrete methods, Rosenbrock method,
Nelder & Mead method, Box’s Complex method, Powell method, Steepest descent
method, Newton's method, conjugate gradient method.
4. Constrained Minimization: Rosen’s gradient projection method for linear 6
constraints, Zoutendijk method of feasible directions for nonlinear constraints,
generalized reduced gradient method for nonlinear constraints.
5. Penalty function methods: Exterior point penalty, Interior point penalty. 4
6. Computer Programs of above methods. Case studies from Engineering and 11
Industry, Use of software packages such as LINDO, LINGO, EXCEL, TORA,
MATLAB
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/Reprint
Bazaraa, M. S., Sherali, H. D. and Shetty, C. M.:”Nonlinear 1993
1. Programming Theory and Algorithms”, 2nd Edition, John Wiley and
Sons.
Belegundu, A. D. and Chandrupatla, T. R. :“Optimization Concepts
2. 2002
and Applications in Engineering”, Pearson Education Pvt. Ltd.
Deb, K.: “Optimization for Engineering Design Algorithms and
3. 1998
Examples”, Prentice Hall of India.
Mohan, C. and Deep, K.: “Optimization Techniques”, New Age India
4 2009
Pvt. Ltd.
Nocedal, J. and Wright, S. J.: “Numerical Optimization”, Springer
5 2000
Series in Operations Research, Springer-Verlag
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-554 Course Title: Reinforcement Learning
2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
:
8. Pre-requisite: Nil
9. Objective: This course aims to understand several reinforcement learning algorithms and their
applications, along with emerging research trends.

10. Details of the Course:

S.No. Contents Contact
Hours
1 Basics of probability and linear algebra, Definition of a stochastic multi-armed 6
bandit, Definition of regret, Achieving sublinear regret, UCB algorithm, KL-
UCB, Thompson Sampling.
2 Markov Decision Problem, policy, and value function, Reward models (infinite 8
discounted, total, finite horizon, and average), Episodic & continuing tasks,
Bellman's optimality operator, and Value iteration & policy iteration
3 The Reinforcement Learning problem, prediction and control problems, 8
Model-based algorithm, Monte Carlo methods for prediction, and Online
implementation of Monte Carlo policy evaluation
4 Bootstrapping; TD(0) algorithm; Convergence of Monte Carlo and batch 6
TD(0) algorithms; Model-free control: Q-learning, Sarsa, Expected Sarsa.
5 n-step returns; TD(λ) algorithm; Need for generalization in practice; Linear 6
function approximation and geometric view; Linear TD(λ).
6 Tile coding; Control with function approximation; Policy search; Policy 8
gradient methods; Experience replay; Fitted Q Iteration; Case studies.
Total 42
11. Suggested Books:
S.No. Name of Authors/Book/Publisher Year of
Publication/Reprint
Sutton, Richard S., and Andrew G. Barto. “Reinforcement learning: An
1 2020
introduction,” First Edition, MIT press
2 Sugiyama, Masashi. “Statistical reinforcement learning: modern 2015
machinelearning approaches,” First Edition, CRC Press
3 Lattimore, T. and C. Szepesvári. “Bandit algorithms,” First Edition, 2020
Cambridge University Press.
4 Boris Belousov, Hany Abdulsamad, Pascal Klink, Simone Parisi, and 2021
Jan Peters “Reinforcement Learning Algorithms: Analysis and
Applications,” First Edition, Springer
5 Alexander Zai and Brandon Brown “Deep Reinforcement Learning in 2020
Action,” First Edition, Manning Publications
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-555 Course Title: Time Series Data Analysis
2. Contact Hours: L:3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil

9. Objective: The objective of this course is to understand and analyze time-series data facilitated by
R programming

10. Details of the Course:

S.No. Contents Contact
Hours
1 Basic Properties of time-series data: Distribution and moments, Stationarity, 4
Autocorrelation, Heteroscedasticity, Normality
2 Autoregressive models and forecasting: AR, ARMA, ARIMA models 4
3 Random walk model: Non-stationarity and unit-root process, Drift and Trend 4
models
4 Regression analysis with time-series data using R programming 5
5 Principal Component Analysis (PCA) and Factor Analysis 5
6 Conditional Heteroscedastic Models: ARCH, GARCH. T-GARCH, BEKK- 6
GARCH
7 Introduction to Non-linear and regime-switching models: Markov regime- 5
switching models, Quantile regression, Contagion models
8 Introduction to Vector Auto-regressive (VAR) models: Impulse Response 5
Function (IRF), Error Correction Models, Co-integration
9 Introduction to Panel data models: Fixed-Effect and Random-Effect models 4
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of
Publication/Reprint

Chris Brooks “Introductory Econometrics for Finance,” Fourth

1 2019
Edition, Cambridge University Press
2 Ruey S. Tsay “Analysis of Time-series data,” Third Edition, Wiley 2014
3 John Fox and Sanford Weisberg “An R Companion to Applied 2018
Regression,” Third Edition, SAGE
4 Yves Croissant and Giovanni Millo “Panel Data Econometrics with 2018
R,” First Edition, Wiley
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-556 Course Title: Introduction to Compressive Sensing

2. Contact hours: L: 3 T: 1 P: 0
3. Examination duration: Theory: 3 Practical: 0
4. Relative weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject area: PEC
8. Pre-requisite: Knowledge of basic concepts in linear algebra, probability and constrained optimization.
9. Objective: The course introduces the basic concepts and mathematics behind compressed sensing and sparse
recovery.

10. Details of the Course:

S.No. Contents Contact

Hours
1. Mathematical Preliminaries: Vector/matrix norms, Orthobasis expansion, 8
Gaussian/Sub- Gaussian random variables and properties, basic concentration
inequalities, basics of convex optimization and constrained optimization
2. Principles of sparse recovery: Unique and stable sparse solutions of 16
underdetermined linear systems, Unique sparse representation and uncertainty
principle, Sensing matrix design, Null-space property (NSP), Mutual coherence
based uniqueness and stable recovery guarantees, Restricted Isometry Property
(RIP), Relationship between RIP and NSP, Johnson-Lindenstrauss lemma,
Sparse recovery with random matrices
3. The compressed sensing problem & connections to sparse recovery: Sparse 3
representation of signals, compressible signals, union of subspaces
4. Sparse recovery methods: Convex optimization algorithms - Basis Pursuit and 10
LASSO, Greedy algorithms - Orthogonal Matching Pursuit (OMP),
Thresholding-based algorithms- Iterative Hard Thresholding (IHT), MAP
estimation-based sparse recovery methods
5. Applications: Sub-Nyquist sampling, Image compression, Image-denoising, 5
Sparse linear regression.
Total 42

11. Suggested Books:

S.No. Name of Authors / Books / Publisher Year of

Publication/Reprint
1. Michael Elad, ``Sparse and Redundant Representations - From Theory 2010
To Applications in Signal & Image Processing”, 2010, Springer
Publications.
2. Simon Foucart and Holger Rauhut, ``A Mathematical Introduction to 2013
Compressive Sensing”, 2013, Birkhauser
3. Yonina Eldar and Gitta Kutyniok, ``Compressed Sensing: Theory and 2012
Applications”, Cambridge University Press
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-557 Course Title: Neuromorphic computing with emerging memories and
architectures

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: This course will teach a student about devices, circuits and architectures for hardware
implementation of neuromorphic systems

10. Details of the Course

S.No. Contents Contact

hours
1. Introduction to Deep learning: Deep Learning fundamentals, Training Deep 6
Architectures, Sigmoid Neurons, Gradient Decent, Feedforward Neural
Networks, Back-propagation, Principal component Analysis and its
interpretations, Singular Value Decomposition, Batch Normalization,
Introduction to Tensor flow.
2. Deep learning Algorithms: Gradient Descent and Back-propagation, Improving 6
deep network, Multi-Layer Neural Networks, The Challenge of Training Deep
Neural Networks, Deep Generative Architectures. Mini-batches, Unstable
Gradients, and Avoiding Over-fitting, Applying deep net theory to code,
Introduction to convolutional neural networks for visual recognition.
3. Advanced Deep Architectures: RNNs, RNNs in practice, LSTMs and GRUs, 6
LSTMs and GRUs in practice, Reinforcement learning, Importance of
unsupervised learning, Auto encoder.
4. Introduction to new trends in computing: Numerical computing, Parallel 8
computing, Cognitive computing, Approximate computing, Near memory and In-
memory computing, Cloud, Fog, and Edge computing, Reconfigurable and
heterogeneous computing.
5. ANN in hardware: General-purpose processors, Digital accelerators, Digital 8
ASIC approach, Optimization on data movement and memory access, Scaling
precision, Leveraging sparsity, FPGA based accelerators, Analog/mixed-signal
accelerators, Neural networks in conventional integrated technology, In/near-
memory computing, Near-sensor computing, Neural network based on emerging
non-volatile memory, Crossbar as a massively parallel engine, Learning in a
crossbar, Case study: An energy-efficient accelerator for adaptive dynamic
programming, Hardware architecture, On-chip memory, Datapath, controller,
Design examples.
 6. Neuromorphic computing with emerging memories: Memristive and CMOS 8
devices for neuromorphic computing, Multi-terminal transistor-like devices based
on strongly correlated metallic oxides for neuromorphic applications, Bipolar
analog memristors as artificial synapses for neuromorphic computing, Robust
memristor networks for neuromorphic computation applications.
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication / Reprint
1. L. Deng and D. Yu, Deep learning: methods and applications, Now 2014
Publishers Inc. (1st edition)
2. M. A. Nielsen, Neural Networks and Deep Learning, MIT Press (1st 2015
edition)
3. I. Goodfellow, Y. Bengio, and A. Courville, Deep learning, MIT Press 2016
(2nd edition)
4. K. H. Mohamed, Neuromorphic Computing and Beyond: parallel, 2021
approximation, near memory, and quantum, Springer (1st edition)
5. Neuromorphic Computing and Beyond by K. S. Mohamed, Springer 2020
(1st edition)
6. J. Suñé, Memristors for Neuromorphic Circuits and Artificial 2020
Intelligence Applications, MDPI AG (1st edition)
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-558 Course Title: Data Stream Mining

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: NIL
9. Objective: To introduce students to the various concepts and techniques in data stream mining

10. Details of the Course

S.No. Contents Contact

hours
1. Introduction to Data Streams: Data stream models, basic streaming methods, 7
data synopsis, sampling, histograms, Wavelets, Discrete Fourier Transform
2. Clustering from Data Streams: Basic concepts, Leader Algorithm, partitioning 7
clustering, hierarchical clustering, grid clustering
3. Frequent Pattern Mining from Data Streams: Search space, landmark 7
windows, mining recent frequent item sets, sequence pattern mining, reservoir
sampling for sequential pattern mining
4. Classification from Data Streams: Decision Trees, VFDT- The base algorithm, 7
extensions to the basic algorithm, exhaustive search, functional tree leaves,
detecting changes
5. Change Detection in Data Streams: Introduction, novelty detection as a one- 7
class classification problem, positive Naïve Bayes, learning new concepts,
approaches based on extreme values, decision structure, frequency distances,
online novelty and drift detection
6. Case Study: Time Series Data Streams – prediction, similarity, symbolic 7
approximation
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication / Reprint
1. Gama, J., “Knowledge Discovery from Data Streams,” 1st Ed., 2010
Chapman and Hall
2. Aggarwal, Charu C., “Data Streams: Models and Algorithms,” 2007
Springer
3. Tan, P.N., Steinbach, M. and Kumar, V., “Introduction to Data 2011
Mining”, Addison Wesley – Pearson.
4. L. Rutkowski, M. Jaworski, P, Duda, “Stream Data Mining: 2020
Algorithms and Their Probabilistic Properties,” 1st Edition, Springer
International Publishing.
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-559 Course Title: Stochastic Processes and their Applications
2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: To introduce concepts stochastic processes and their applications.

10. Details of the Course

S.No. Contents Contact

hours
1. Review of Probability: Probability measure, Borel-Cantelli lemma, multivariate 7
random variable, Doob-Dynkin lemma, expectation, joint distribution and joint
density functions, conditional expectation and its properties, conditional
distribution and conditional density functions, independence of random variables,
Markov inequality, Chebyshev inequality, convergence of random variables, law
of large numbers, related applications and simulations.
2. Discrete time Markov chain: Definition and construction, transition probability 11
matrix, higher order transition probabilities, Chapman-Kolmogorov equation,
dissection principle, classification of states, periodicity, solidarity properties,
canonical decomposition, absorption probabilities, invariant measure and
stationary distribution, limit distributions, renewal process, branching process,
related applications and simulations.
3. Continuous time Markov chain: Definition and construction, examples (pure 11
birth process, birth-death process, uniformizable chain, etc.), stability and
explosions, Markov property, dissection, backward and forward equations,
generator, Chapman-Kolmogorov equation, stationary and limiting distributions,
invariant measure, Laplace transform method, generating function technique, Point
process, Poisson process, compound Poisson process, renewal process, Branching
process, related applications and simulations.
4 Brownian Motion: Definition and construction (via random walk and Brownian 6
bridge approximations), sample path properties, Brownian motion with drift,
Ornstein-Uhlenbeck process, related applications and simulations.
5 Martingales: Filtration, stopping time, discrete time martingales with examples, 7
optional stopping theorem, Doob’s up-crossing inequality, Doob’s convergence
theorem, Doob’s decomposition theorem, continuous time martingales with
examples, related applications and simulations.
Total 42
11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/ Reprint
1. Zdzislaw Brzezniak and Tomasz Zastawniak, Basic stochastic 2005/7th
processes, Springer-Verlag London
2. Sindney Resnick, Adventures in stochastic processes, Birkhäuser 2005/4th
Boston
3. Paul Glasserman, Monte Carlo Methods in Financial Engineering, 2003/1st
Springer
4. Peter W. Glynn and Søren Asmussen, Stochastic Simulation: 2007/1st
Algorithms and Analysis, Springer
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-560 Course Title: Artificial Intelligence for Decision Making
2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: To introduce the concept of AI integrated decision making systems to the students.
10. Details of the Course:

S.No. Contents Contact

Hours
1 Introduction: Review of decision making process in optimization and 8
operations research models; overview of machine learning algo- rithms;
ranking methods.
2 Network flow models and their integration with AI algorithms: 10
Transportation and transshipment models; travelling salesman problem;
vehicle routing; project management; integration of these models with
ANN, Fuzzy logic, Genetic Algorithms.
3 Multi criteria decision making (MCDM): MCDM methods and their 10
integration with fuzzy logic, ANN; Integration of MCDM methods with
dimensionality reduction techniques like Principle Component Analysis,
Singular Value Decomposition and page rank algorithms.
4 AI integrated inventory models: Basic inventory models; demand 10
prediction for inventory management; reinforcement learning systems for
full inventory management; AI algorithms for prediction and forecasting of
inventory.
5 Implementation: Implementation of the above models in MATLAB/ Python. 4
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/Reprint
1. F.S. Hillier and G.J. Liberman “Introduction to Operations 2001
Research”. Tata McGraw Hill Education Private Limited.
2. H.A. Taha, “Operations Research, an Introduction”, Pearson 2007
3. Michael Carter, Camille C. Price and Ghaith Rabadi 2018
“Operations Research, A Practical Introduction”, CRC press
4. Adiel Teixeira de Almeida, Emel Aktas, Sarah Ben Amor, João 2020
Luis de Miranda “Advanced Studies in Multi-Criteria Decision
Making“, CRC Press.
5. Gregory S. Parnel, Terry A. Bresnick, Steven N. Tani, Eric R. 2013
Johnson “Handbook of Decision Analysis”, Wiley.
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-561 Course Title: AI for Earth Observation

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs): Theory: 3 Practical: 0
4. Relative Weight: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Spring 7. Subject Area: PEC
8. Pre-requisite: Machine Learning
9. Objective of Course: Earth Observation is a key application area of AI. The objective of the
course is to understand the application of AI in Earth Observation.

10. Details of the Course:

Contact
S.No. Contents
Hours
Physical Fundamentals of Earth Observation: Brief History of the
Development of Earth Observation Sensors, Physical Properties of Electro-
Magnetic Waves, Introduction to Electro-Magnetic Spectrum and Its Use in Earth
1 Observation, Hyperspectral Remote Sensing
Sensors and Data: Types of Resolutions, Types of Sensors: Optical, Microwave, 6
Non- Imaging Sensors, UAV, Satellite Observation Geometries, Atmospheric
Emissions
Data Science Pipeline in Earth Observation: Data Discovery and Organization
of Data; Accessing Data; Exploratory Data Analysis and Visualization; Creation
2 of Labels/Training Data; Analysis and Knowledge Discovery [Application of ML 6
& DL]; AccuracyAssessment
Analysis and Knowledge Discovery using SVM, Random Forest, SOM, CNN,
RNN, LSTM, GANs with:
(a) Earth Observation Image Classification
3 (b) Automatic Target/Object Detection and Classification 10
(c) Time Series Analysis
(d) Disaster Monitoring
(e) Agriculture; Infrastructure; Weather and Space Weather
4 Transfer Learning using AI models in Earth Observation 4
5 EO Data Requirements: Database Techniques for Storing EO Data and Training 6
Data; Relational Geospatial Big Data Systems
6 Review of Current Research and Practices in AI for EO 4
7 Mini Project on the Application of AI for Analysing a Specific Domainin EO 6
Total 42
List of Tutorials:

Tutorial 1: Access to different EO sensors, open EO datasets from different space agencies
Tutorial 2: Data discovery and accessing data using API, exploratory data analysis and visualization of
EO data.
Tutorial 3: Application of ML models for EO data analysis and knowledge discovery.
Tutorial 4: Application of DL models for EO data analysis and knowledge discovery.
Tutorial 5: Transfer Learning of DL models in EO.
Tutorial 6: EO Scalable Data formats and geospatial big-data systems

11. Suggested Books

Year of
S.No. Name of Authors/ Books/ Publisher
Publication/Reprint
Thenkabail, P.S. “Remotely Sensed Data Characterization,
1 Classification, and Accuracies”: Three Volumes, First Edition, 2015, 2016
CRC Press
Goodfellow, I., Courville, A., Bengio, Y. “Deep Learning”, 2016, MIT
2 2017
Press
Aurélien Géron. “Hands-On Machine Learning with Scikit-Learn, Keras,
3 and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent 2019
Systems”, Second Edition, 2019, O’Reilly

12. Suggested software/computer languages to be used in the course

S.No. Name of software

1 Python and Jupyter Notebooks; TensorFlow; PyTorch
2 Google Earth Engine
3 QGIS
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-562 Course Title: AI for Investment

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs): Theory: 3 Practical: 0
4. Relative Weight: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective of Course: The objective of this course is to understand the application of Artificial
Intelligence and Machine Learning techniques in financial markets, trading, and asset management.

10. Details of the Course:

Contact
S.No. Contents
Hours
1 Introduction to financial markets and market microstructure 4
2 Introduction to risk-return framework 4
3 Introduction to asset management and portfolio optimization 4
4 Market efficiency and behavioral finance 4
5 Prediction in Financial markets using AI and machine learning models, AI and 6
machine learning in Trading execution and portfolio management
6 Credit scoring and credit modeling with non-linear machine learning models and 4
deep learning
7 Model risk management and stress testing 4
8 Robo advisory, social and quantitative investing 5
9 Machine learning for asset management 4
10 AI and machine learning in regulatory compliance and supervision 3
Total 42

11. Suggested Books:

Year of
S.No. Name of Authors/ Books/ Publisher
Publication/Reprint
M. Dixon, I Halperin, and P. Bilokon “Machine Learning in
1 2020
Finance,” First Edition, Springer
2 Marcos Lopez “Advances in Financial Machine Learning,” First 2018
Edition, Wiley
3 Marcos Lopez “Machine Learning for Asset Managers,” First 2020
Edition,Cambridge University Press
4 Stefan Jansen “Machine Learning for Algorithmic Trading,” Second
2020
Edition, Packt
5 Elton and Gruber, “Modern Portfolio Theory,” Ninth Edition, Wiley 2014
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-563 Course Title: Applications of AI in Physics

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE:0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Machine Learning
9. Objective: To enable the students to become an application engineer to apply AI tools to solve
problems in cutting edge physics research.

10. Details of Course:

S. No. Contents Contact

Hours
1. Introduction to big data sets in Physics: Overview of different areas of physics 8
and highlight areas where AI and ML is becoming an important tool of research;
example of big data sets from physics; characterize the data sets from machine
learning and AI point of view; why is Machine Learning difficult -setting up a
physics problem as a ML task.
2. Statistical physics ideas relevant for AI algorithms – Entropy, information, cost 6
function, and minimization from a physics point of view.
3. Application of AI tools to simple physics example - Ising model of Physics; 8
application of selected supervised and unsupervised ML algorithms to Ising
model. Physics-inspired algorithms for better machine learning.
4. Application of AI tools to Condensed Matter Physics - Introduction to the area 8
of research, Application of ML and AI tools to selected examples.
5. Application of AI tools to Radiation Measurement and Modelling - 6
Introduction to Radiation models, Measurement methods, and Application of
ML and AI tools to selected examples.
6. Sensor Designs and deep neural network: Plasmonic sensors modelling 4
Total 40
11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/ Reprint

Pankaj Mehta, Marin Bukov, Ching-Hao Wang, Alexandre G.R.

1 Day, Clint Richardson, Charles K. Fisher, David J. Schwab, A 2019
high-bias, low-variance introduction to Machine Learning for
physicists, by, Physics Reports 810 (2019)
R. Feynman, R. Leighton, and M. Sands, The Feynman
2 Lectures on Physics: The New Millennium Edition: Mainly 1963
Mechanics, Radiation, andHeat, v. 1, ISBN 9780465040858
3 M. Nakhostin, Signal Processing for Radiation Detectors,Wiley, 2017
ISBN: 978- 1-119-41022-
Oliveira, L.C., Lima, A.M.N., Thirstrup, C., Neff, H.F., Surface
4 Plasmon Resonance Sensors, A Materials Guide to Design, 2019
Characterization, Optimization, and Usage, Springer International
Publishing, ISBN 978-3- 030-17485-9
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-564 Course Title: Medical Physics for AI

2. Contact Hours: L: 3 T: 0 P: 2
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 10-25 PRS: 25 MTE: 15-25 ETE: 30-40 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Machine Learning and Python Programming
9. Objective: To provide various applications of artificial intelligence in Medical physics.

10. Details of the Course

S.No. Contents Contact

Hours
1. Introduction to Radiation Modalities: Basics of Imaging Modalities, X-Ray 8
Radiography, X-Ray CT, Ultrasonography, OCT, OCT Angiography, PET &
SPECT, Magnetic Resonance Imaging, miscellaneous biomedical devices.
2. Human Anatomy for AI-aided Diagnostics: General Anatomy, Bones & 8
Joints, Muscle, Respiratory system, Digestive System, Cardiovascular system,
Nervous system, Sense organs.
3. Functional Imaging Analysis: Feature Selection, ML/DL model building, data 8
preparation, model training, and model validation for various Modality, logistic
regression & statistical inference, difference between biological, Experimental
and clinical data. Limitations of AI
4. Radiotherapy and AI: Brief introduction to diseases, computer‐aided 6
detection, classification, and diagnosis in radiology and auto‐contouring,
treatment planning, response modeling (radiomics), image guidance, motion
tracking, and quality assurance in radiation oncology.
5. AI in Cardiology: Brief introduction to diseases, CMR, Heart, Lungs, Head and 6
Neck, RIC
6. Physiological Parameters and AI: Data Analysis using EEG, ECG, SpO2 6
content.
Total 42
11. List of Experiments:

1. Read the DIACOM format from industrial / commercial MRI, CT and SPECT Machines.
2. Identify the body part from given image and categorize into anatomical system.
3. Identify the time series images to synchronize the random images according to human
anatomy.
4. Identify the anatomical and pathological abnormalities from a given image set.
5. Identify the physiological abnormalities from a given data set.
6. Manually segment MRI and CT Images of Heart, lungs and digestive system using Semi-
automatic soft tools.
7. Manually segment OCT Images of Eye using Semi-automatic soft tools.
8. Segmentation of medical images using CNN.
9. Identification of breathing pattern from ECG using CNN.
10. Categorization of sleeping pattern from EEG using CNN.
11. Deep Learning model and CT / OCT Image segmentation.

12. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/Reprint

1 Guyton and Hall Textbook of Medical Physiology. Second South Asia

Edition 2019
2 Classification Techniques for Medical Image Analysis and 1st Edition 2019
ComputerAided Diagnosis, Academic Press.
3 Pattern Classification of Medical Images: Computer Aided 2017
Diagnosis, Springer.
4 Deep Learning in Medical Image Analysis: Challenges and 2020
Applications, Springer.
5 Atam P. Dhawan, Medical Image Analysis. Wiley-IEEE Press. 2011
6 Adam Bohr, Artificial Intelligence in Healthcare, Academic Press 2020
ISBN 978-0-12-818438-7
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-565 Course Title: Computer Vision

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: To provide knowledge about various computer vision techniques and applications of
machine learning in Computer Vision.

10. Details of the Course:

S.No. Contents Contact

Hours
1. Image formation and camera calibration: Introduction to computer vision, 8
geometric camera models, orthographic and perspective projections, weak-
perspective projection, intrinsic and extrinsic camera parameters, linear and
nonlinear approaches of camera calibration
2. Feature detection and matching: Edge detection, interest points and 6
corners, local image features, feature matching and Hough transform, model
fitting and RANSAC, scale invariant feature matching
3. Stereo Vision: Stereo camera geometry and epipolar constraints, essential 12
and fundamental matrix, image rectification, local methods for stereo
matching: correlation and multi-scale approaches, global methods for stereo
matching: order constraints and dynamic programming, smoothness and
graph based energy minimization, optical flow
4. ML in Computer Vision: Image Recognition; Tracking; Pre-trained CNN 10
models in computer Vision; Open-CV; Applications of machine learning in
computer vision
5. Structure from motion: Camera self-calibration, Euclidean structure and 6
motion from two images, Euclidean structure and motion from multiple
images, structure and motion from weak-perspective and multiple cameras
Total 42
11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/Reprint
1. Forsyth, D. A. and Ponce, J., "Computer Vision: A Modern 2011
Approach",Prentice Hall, 2nd Ed.
2. Szeliki, R., "Computer Vision: Algorithms and Applications", 2011
Springer
3. Hartley, R. and Zisserman, A., "Multiple View Geometry in 2003
Computer Vision", Cambridge University Press
4. Gonzalez, R. C. and Woods, R. E., "Digital Image Processing", 2009
Prentice Hall, 3rd Ed.
5. Trucco, E. and Verri, A., "Introductory Techniques for 3-D 1998
ComputerVision", Prentice Hall
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-566 Course Title: Game theory

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: The objective of this course is to understand algorithmic game theory and its
applications using AI and machine learning techniques.

10. Details of the Course:

S.No. Contents Contact
Hours
1 Introduction to Game Theory, Dominant Strategy Equilibria, Pure Strategy 6
Nash Equilibria, computing Nash equilibrium
2 Mixed Strategy Nash Equilibria, Maxmin and Minmax Values, Matrix Games 6
3 Correlated Strategies and Correlated Equilibrium, Nash Bargaining Problem, 10
Coalitional Games with Transferable Utility, The Core, Shapley Value,
Nucleolus
4 Sequential learning in games, multi-agent learning using game theory 6
5 Introduction to Mechanism Design, Arrows Impossibility theorem, Gibbard- 8
Satterthwaite Theorem, Mechanisms with Money
6 Myerson’s Lemma and VCG Mechanism 6
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of
Publication/Reprint
Martin J. Osborne “An Introduction to Game Theory,” First Edition,
1 2003
Oxford University Press.
Y. Narahari “Game theory and mechanism design,” First Edition,
2 2014
World Scientific.
Noam Nisan, Tim Roughgarden, Éva Tardos, Vijay V. Vazirani.
3
“Algorithmic Game Theory,” First Edition, Cambridge University 2007
Press
Ivan Pastine, Tuvana Pastine, and Tom Humberstone “Introducing
4 2017
Game Theory: A Graphic Guide,” First Edition, Icon Books Ltd
Michael Maschler, Eilon Solan, Shmuel Zamir “Game Theory,”
5 2020
Second Edition, Cambridge University Press
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Department of Computer Science and Engineering

1. Subject Code: CSN-527 Course Title: Internet of Things

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Knowledge of computer networks
9. Objective: To impart the know-how of Internet of Things and their applications, architectures and
protocols, building IoT applications/systems, securing the IoT systems, and their recent advances.

10. Details of the Course

S.No. Contents Contact

hours
1. Basic concepts revisited: Introduction to sensing & actuating, Basic networking, 7
Wireless networks, Wireless sensor networks (WSN), Communication protocols,
and other enabling technologies, IoT standards, Data storage & management
issues and approaches, Cloud computing, Key challenges, research, and future
directions of IoT, and security & privacy issues.
2. Embedded Systems: Hardware and software of IoT, Microcontrollers, 6
Understanding and programming Arduino, Raspberry Pi, NodeMCU, Lora, etc.
Integrating microcontrollers with sensors and actuators, Building the IoT
applications with any microcontroller.
3. IoT Architectures and Protocols: Layers of communication, Architectures: 9
State-of-the-art, IoT architecture reference models, Different views of IoT
architectures and frameworks design, Protocols: Application protocols, Service
discovery protocols, Infrastructure protocols, and other protocols. Understanding
various types of protocols like HTTP, MQTT, CoAP, AMQP, 6LoWPAN, etc.
Cross-layer implementations, and Data dissemination.
4. Support Technologies for IoT: Big Data, Data Analytics, Artificial Intelligence, 8
Mobile, Cloud, Software defined networks, 5G, and Fog/Edge computing. IoT
integration with recent technologies. State-of-the-art. Design goals, challenges,
and components.
5. Cyber Physical Systems: Industry 4.0, Society 5.0, Design & use cases, 6
Development, and implementation insights some examples like smart cities, smart
homes, smart grids, smart agriculture, smart healthcare, smart transportation,
smart manufacturing, and other smart systems. State-of-the-art. Conceptualizing
the new IoT-based smart systems using a case study.
6. IoT Security & Privacy: –, IoT Security and Privacy issues and challenges, Risks 6
involved with IoT infrastructures, Trust in IoT platforms and other integrating
technologies, Data aggregation, storage, retrieval, and other management issues
including fault tolerance, interoperability, security, and privacy, Cyber-physical-
systems and their security and privacy, Mitigation approaches.
Total 42
11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication / Reprint
1. Edited by: Buyya, Rajkumar, and Amir Vahid Dastjerdi, Internet 2016
of Things: Principles and paradigms. Elsevier/Morgan Kaufmann
2. Bahga, Arshdeep; Madisetti, Vijay, Internet of Things (A Hands- 2014
on-Approach), AbeBooks.com
3. Sohraby, Kazem, Daniel Minoli, and Taieb Znati. Wireless sensor 2007
networks: technology, protocols, and applications. John Wiley &
Sons
4. Marinescu, Dan C., Cloud computing: theory and practice. 2017
Elsevier/ Morgan Kaufmann
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-567 Course Title: Introduction to Materials Informatics

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: The course will introduce the students to the applications of data analysis and
machine learning methods to the materials science problems. The course will provide an
introduction to basic informatics and then focus on their application in materials synthesis,
structural design and propertyoptimization.

10. Details of the Course

S.No. Contents Contact

Hours
1 Introduction to Materials Informatics: History of materials 8
development and need for new approaches, Multiscale materials
modelling, need for data driven modelling, accelerated materials
discovery and development, Quantitative structure-processing- property-
performance relationships, knowledge discovery workflow for materials
informatics, materials data science – structured and unstructured data,
data mining, crystallography data base, Materials Genome, different
sets of descriptors, nuts and bolts of materials informatics.
2 Optimization - Calibration: gradient based optimization, non- gradient 8
based optimization, multi objective genetic algorithms (MOGA),
Optimization of a multivariate model, applications to materials synthesis,
processing, and transport phenomena.
3 Predictive Modelling: supervised learning, regression methods, 8
classification methods, surrogate based optimization, prediction of
material properties such as fatigue life, creep life.
4 Descriptive Modelling: Unsupervised learning, clustering analysis, 8
clustering algorithms. Case studies: Estimation of microstrain, residual
stress from diffraction, classification of materials based on physical
properties.
5 Limitations and Remedies: Problem of small datasets in materials 6
science, Data dimensionality reduction – principal component analysis,
applications to 4D diffraction, spectroscopic data sets, high-throughput
computational modelling of materials.
6 Materials Selection for Engineering Design: Systematic selection 4
methods, trade-off analysis, vectors for materials development
Total 42
11. Suggested Books:

Year of
S.No. Name of Authors /Books/ Publisher Publication/ Reprint
1 Informatics for Materials Science and Engineering, Edited 2013
by Krishna Rajan, 1st edition, Butterworth-Heinemann, ISBN:
978-0- 123-94399-6
2 Materials Informatics: Methods, Tools, and Applications, 2019
Edited by Olexandr Isayev, Alexander Tropsha and Stefano
Curtarolo, 1st edition, Willey, ISBN: 978-3-527-34121-4
3 S.R. Kalidindi, Hierarchical Materials Informatics, 1st edition, 2015
Butterworth-Heinemann, ISBN: 978-0-124-10394-8
4 Nanoinformatics, Edited by Isao Tonaka, 1st edition, Springer 2018
Nature, ISBN: 978-9-811-07616-9 (Open access eBook)
5 Information Science for Materials Discovery and Design, 2016
Edited by Turab Lookman, Francis Alexander and Krishna
Rajan, 1st edition, Springer, ISBN: 978-3-319-23870-8
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Department of Computer Science and Engineering

1. Subject Code: CSN-519 Course Title: Social Network Analysis

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Spring 7. Subject Area: PEC
8. Pre-requisite: Knowledge of computer networks
9. Objective: To introduce the basic notions used for social network analysis.
10. Details of the Course

S.No. Contents Contact

hours
1. Social Network Analysis: Preliminaries and definitions, Erdos Number Project, 4
Centrality measures, Balance and Homophily.
2. Random graph models: Random graphs and alternative models, Models of 4
network growth, Navigation in social Networks
3. Network topology anddiffusion, Contagion in Networks, Complex contagion, 4
Percolation and information, Epidemics and information cascades
4. Cohesive subgroups, Multidimensional Scaling, Structural equivalence, Roles 6
and positions, Ego networks, Weak ties, Structural holes
5. Small world experiments, Small world models, Origins of small world, Heavy 6
tails, Small Diameter, Clustering of connectivity
6. The Erdos Renyi Model, Clustering Models, Preferential Attachment 6
7. Navigation in Networks Revisited, Important vertices and page rank algorithm, 6
Towards rational dynamics in networks, Basics of game theory
8. Coloring and consensus, biased voting, network formation games, network 6
structure and equilibrium, behavioral experiments, Spatial and agent-based
models
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication / Reprint
1. Wasserman, Stanley, and Joseph Galaskiewicz. Advances in social 1994
network analysis: Research in the social and behavioral sciences.
Sage
2. Knoke, David, and Song Yang. Social network analysis. Sage 2019
Publications
3. Carrington, Peter J., John Scott, and Stanley Wasserman, eds. 2005
Models and methods in social network analysis. Vol. 28.
Cambridge university press.
4. Liu, Bing. "Social network analysis." In Web data mining, pp. 269- 2011
309. Springer, Berlin, Heidelberg
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE
NAME OF DEPARTMENT/CENTRE: Department of Electronics and Communication Engineering

1. Subject Code: ECN-526 Course Title: Statistical Machine Learning for Variation-
Aware Electronic Device and Circuit Simulation
2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Spring 7. Subject Area: PEC
8. Pre-requisite: Knowledge of basic concepts in probability and statistics
9. Objective: To familiarize students with the fundamental concepts, techniques and
algorithms needed to perform stochastic simulation and uncertainty quantification of
electronic devices, circuits and systems.

10. Details of the Course

S.No. Contents Contact

hours
1. Introduction: Introduction to stochastic modeling of general systems, 2
key differences between stochastic simulation and classical deterministic
simulation. The need for uncertainty quantification in general device, circuit,
and system simulation.
2. Introduction to Random Variables: Discrete and continuous random 3
variables: distribution and density functions, conditional distributions and
expectations, functions of random variables, statistical moments, sequence of
random variables, central limit theorem, Gaussian and non-Gaussian
correlation among random variables
3. Random Sampling Techniques: Utilization of random sampling techniques 5
for statistical analysis such as Monte Carlo, quasi-Monte Carlo, Latin
hypercube sampling, analysis of computational complexity and convergence
rate of different random sampling techniques
4. Statistical Machine Learning - Generalized Polynomial Chaos (PC) 6
Theory: Basic foundation of polynomial chaos, generalization of polynomial
chaos for different known distributions, Wiener-Askey scheme of
polynomials, generation of orthonormal basis functions using three-term
recurrence series and Gram-Schmidt algorithm, training of polynomial chaos
metamodels using quadrature techniques and least-squares linear regression.
Deployment of PC theory for calculating statistical moments and density
functions in linear and nonlinear VLSI as well as RF/microwave devices,
circuits, and systems via test cases and illustrative examples.
5. Correlations in PC Theory: Considering uncorrelated, Gaussian correlated, 5
and non-Gaussian (mixed Gaussian model) correlated parametric variations.
6. Advanced PC theory: Complexity analysis of PC theory and techniques: 13
limitations of curse of dimensionality in PC theory, emphasis on
sensitivity analysis-based dimension reduction, active subspaces,
sliced inverse regression compressed sensing, partial least-squares
algorithm, and multi-fidelity methods.
 7. Inverse Problems: Bayes rule, Bayesian formulation of inverse 8
problems, prior and posterior distributions, calculation of maximum
likelihood function using PC theory. Applications into inverse
uncertainty quantification in linear/nonlinear devices, circuits and systems
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/ Reprint
1. D. Xiu, “Numerical Methods for Stochastic Computations: 2010
A Spectral Method Approach,” New Jersey: Princeton
University Press
2. D. Dubois and H. Prade, “Possibility Theory: An Approach 1988
to Computerized Processing of Uncertainty,” vol. 2, New York,
NY: Plenum Press
3. K. C. Gupta and Q. J. Zhang, “Neural Networks for RF and 2000
Microwave Design,” Arctech House
4. A. Papoulis and S. Pillai, "Probability, Random Variables and 2017
Stochastic 2017 Processes", 4th Edn., Mc Graw Hill.
5. R. Shen, S. X.-D. Tan, and H. Yu, Statistical Performance 2012
Analysis and Modeling of Nanometer VLSI. New York,
NY:Springer
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-568 Course Title: ML and AI Applications in Earth Sciences

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Good foundation in Mathematics and Physics with specific exposure in Numerical
Methods. Understanding of fundamental principles of Geology and Geophysics would be preferable.

9. Objective: To make the participants familiar with tools and techniques in Earth Sciences and the use of
Machine Learning and Artificial Intelligence for optimizing the workflows for more accurate prediction
of events and properties of the subsurface.

10. Details of the Course:

S.No. Contents Contact

Hours
1 Familiarization with Major Domains and Data Types in Earth Sciences: 4
Earthquake Seismology, Engineering Geology and Rock Mechanics, Reservoir
Characterization, Paleontology
2 General Introduction to ML and AI in Earth Sciences: 6
ML and statistical pattern recognition: Supervised learning (generative/ descriptive
learning, parametric/ non-parametric learning, neural networks, Support vector
machines), Unsupervised learning (clustering, dimensionality reduction, kernel
methods); time series modelling, linear regression, regularization, linear classifiers,
ensemble methods, neural networks, model selection and evaluation, scalable
algorithms for big data, and data ethics.
Data science: Extreme value statistics, multi-variate analysis, factor analysis,
compositional data analysis, spatial information aggregation models, spatial estimation,
geo-statistical simulation, treating data of different scales of observation, spatio-
temporal modelling (geo-statistics).
3 Automating Data Mining and Analysis in Seismology: Basics of earthquake 6
detection and phase picking using short-term average (STA)/long-term average (LTA);
detection using waveform similarity: Network Matched Filtering/template matching,
Fingerprint And Similarity Thresholding (FAST). Associating seismic phases across all
stations using deep-learning techniques and combining the ones have the same origin
source (PhaseLink). Generic workflow of data collection, preprocessing, model
training, model evaluation, and production. Applications of ML in ground motion
synthesis, and future directions.
4 Classification of Earthquake Sources: Using supervised learning for classifying 4
earthquakes and finding their occurrence mechanism. Training dataset (waveforms) on
different kinds of sources: earthquake, glacial, volcanic, landslide, explosion, etc. A
brief discussion on seismic sources and radiation pattern of emerging waves.
5 Deep learning (DL) based Seismic Inversion: Theory of Seismic Inversion, 4
Convolutional neural network (CNN) and fully connected network (FCN) architectures,
Performance evaluation, Geophysical inversion versus ML, their applications to
reflectivity inversion in seismic, Numerical examples.
 6 Automation in 3D Reservoir Property Prediction: Data Mining, Automated 4
Petrophysics, Statistical and Regression Methods for Elastic Property Prediction, ML
and AI application in Geostatistics, Convoluted Neural Networks for Seismic
Interpretation, Deep Learning for Impedance Inversion and Porosity Prediction.
7 Data-Driven Analytics in Shale Resources: Concepts of shale as source-reservoir- 4
seal, Modeling Production from Shale, Shale Analytics, Decline Curve Analysis, Shale
Production Optimization Technology (SPOT), Numerical Simulation and Smart Proxy
8 Machine learning Applications in Engineering Geology and Rock Mechanics: ML 6
in rock mass characterization, Rock Mass Rating, Slope Mass Rating, Q-System,
Engineering properties of rock and various rock engineering applications, AI in
Landslides study.
9 Separation and Taxonomic Identification of Microfossil: 3D object recognition and 4
segmentation applied to X-ray MicroCT images; Testing different algorithms for
identifying and localizing individual microfossils in rock samples: Automated
Computer Vision, Deep learning-based CNN semantic, and other segmentation
architectures.
Total 42

11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publications/ Reprint
1 Patrick Wong, Fred Aminzadeh, and Masoud Nikravesh, Soft 2002
Computing for Reservoir Characterization and Modeling, Springer-
Verlag Berlin Heidelberg GmbH
2 William Sandham & Miles Leggett, Geophysical Applications of 2003
Artificial Neural Network and Fuzzy Logic, Springer
3 C. Cranganu, H. Luchian, M. E. Breaban, Artificial Intelligent 2015
Approached in Petroleum Geosciences, Springer
4 Shahab D. Mohaghegh , Data-Driven Analytics in Unconventional 2017
Resources, Springer
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Department of Electrical Engineering

1. Subject Code: EEN-581 Course Title: Intelligent Control Techniques

2. Contact Hours: L: 3 T: 0 P: 2
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 10-25 PRS: 25 MTE: 15-25 ETE: 30-40 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Control Systems
9. Objective: To introduce soft computing and intelligent control techniques and to apply these techniques
to solve real-world modelling and control problems.

10. Details of the Course

S.No. Contents Contact

hours
1. Fuzzy Logic Systems: Fuzzy sets, operations on fuzzy sets, fuzzy relations,
operations on fuzzy relation, linguistic variables, fuzzy if then rules, compositional 6
rule of inference, fuzzy reasoning.
2. Fuzzy Logic Control: Basic concept of fuzzy logic control, reasoning with an FLC,
relationship to PI, PD and PID control,design of FLC: determination of linguistic 6
values, construction of knowledge base, inference engine, tuning, fuzzification and
defuzzification, Mamdani type models, Takagi-Sugeno-Kang (TSK) fuzzy models.
3. Artificial Neural Networks: Perceptrons, perceptron training rule, gradient descent
rule, multilayer networks and backpropagation algorithm, convergence and local 12
minima, regularization methods, radial basis function networks, alternative error
functions, alternative error minimization procedures, recurrent networks, extreme
learning machines, unsupervised networks.
4. Neural Networks for feedback Control: Identification of system models using
neural networks, Model predictive control, feedback linearization and model 8
reference control using neural networks, Neural Network Reinforcement Learning
Controller, Adaptive Reinforcement Learning Using Fuzzy Logic Critic, Optimal
Control Using NN.
5. Hybrid algorithms: Neuro fuzzy systems, ANFIS and extreme learning ANFIS,
derivative free optimization methods, genetic algorithm, particle swarm 8
optimization, solution of typical control problems using derivative free
optimization
Total 40
11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication / Reprint
1. Christopher M. Bishop, Neural Networks for Pattern Recognition”, 1995
Oxford University Press, New York
2. S. Haykin, Neural Networks and Learning Machines, Prentice Hall 2009
3. Driankov, Hellendoorn, Reinfrank, An Introduction to Fuzzy Control, 1993
Narosa Publishing House
4. Timothy J. Ross., Fuzzy Logic with Engineering Applications, John 2011
Wiley and Sons
5. SR Jang, CT Sun, E Mizutani, Neuro-fuzzy and soft computing: 2004
a computational approach to learning and machine intelligence,
Prentice-Hall of India
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-569 Course Title: Applications of AI in Biology

2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: The course provides introduction to AI, Machine Learning and Deep learning algorithms,
hands-on experience using Python and exposure to applications in genomics, medicine, biological
and biomedical image analysis and in general computational biology and bioinformatics by discussion
around published research.

10. Details of the Course

S.No. Contents Contact

hours
1. Relevance of ML in Biology and Medicine; Glimps of AI applications in Biology 4
and Medicine; Handling biological and bioinformatics data; tools for data
handling;
2. Supervised Machine Learning applications in Biology and Medicine; Regression 6
models based examples in Biology
3. Applications of Decision trees, Random Forest, Support Vector Machines models 6
in biology and medicine.
4. Applications of Clustering Methods (k-means, Hierarchical, DBSCAN). 6
Dimension Reduction: PCA, t-SNE. in Biology using research publications.
5. Probabilistic Models, GANs, Hidden Markov Models, EM Algorithm. Paper 6
examples using various algorithms
6. Some well-known fully connected and deep networks and their use in Biological 8
applications; case studies; Explore different ways Deep Learning is used in Biology
through papers
7. Some related case studies; Discussing/Presenting papers that that uses AI/ML/DL 6
specifically related to biological applications
Total 42
11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication/Reprint
1 Kevin Murphy, “Machine Learning: A Probabilistic Approach” 1st 2012, 2021
Edition (The MIT Press)
2 Pierre Baldi and Soren Brunak, “Bioinformatics: The Machine 2001
Learning Approach” 2nd Edition (The MIT Press)
3 Tom M. Mitchell, “Machine Learning” (McGraw-Hill) 1997
4 Ian Good fellow, Yoshua Bengio and Aaron Courville, Online book
deeplearningbook.org(MIT Press)
5 Christopher M. Bishop “Pattern Recognition and Machine Learning” 2006
Springer
6 The Elements of Statistical Learning: Data Mining, Inference, and 2009
Prediction. T.Hastie, R. Tibshirani, J. Friedman, 2nd Edition
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-570 Course Title: VLSI architectures for AI in CMOS Technology
2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: NIL
9. Objective: This course will teach the students about efficient implementation of computation intensive
AI algorithms and operations using VLSI devices.

10. Details of the Course

S.No. Contents Contact

hours
1. Algorithms for fast addition: Basic addition and counting, Bit-serial and ripple- 6
carry adders, Manchester carry chains and adders, Carry-look-ahead adders, Carry
determination as prefix computation, Alternative parallel prefix networks, VLSI
implementation aspects, Variations in fast adders, Simple carry-skip and Carry-
select adders, Hybrid adder designs, Optimizations in fast adders, Multi-operand
addition, Wallace and Dadda trees.
2. High speed multiplication: Basic multiplication schemes, Shift/add multiplication 6
algorithms, Programmed multiplication, Basic hardware multipliers, Multiplication
of signed numbers, Multiplication by constants, Preview of fast multipliers, High-
radix multipliers, Modified Booth's recoding, Tree and array multipliers, Variations
in multipliers.
3. Real Arithmetic: Representing the real numbers, floating-point arithmetic, The 8
ANSI/IEEE floating point standard, Floating-point arithmetic operations,
Rounding schemes, Logarithmic number systems, Floating-point adders, Barrel-
shifter design, Leading-zeros/ones counting, Floating-point multipliers, Floating-
point dividers, Arithmetic Errors and error control.
4. Implementation Topics: Computing algorithms, Exponentiation, Approximating 6
functions, Merged arithmetic, Arithmetic by table lookup, Tradeoffs in cost, speed,
and accuracy. High-throughput arithmetic, Low-power arithmetic, Fault-tolerant
arithmetic, Impact of hardware technology.
5. VLSI architectures: Analog VLSI neural learning circuits, An analog CMOS 8
implementation of Kohonen network with learning capability, Backpropagation
learning algorithms for analog VLSI implementation, Analog implementation of
the Boltzmann machine with programmable learning algorithms, VLSI design of
the minimum entropy neuron.
6. VLSI Designs: VLSI design of a 3-D highly parallel message-passing architecture, 8
A dataflow architecture for AI, Processing in-memory design, COLIBRI:
Coprocessor for LISP based on RISC.
Total 42
11. Suggested Books:

S.No. Name of Authors/Book/Publisher Year of

Publication / Reprint
1. B. Parhami, Computer Arithmetic: Algorithms and Hardware 2010
Designs, Oxford University Press, New York (2nd edition)
2. I. Koren, Computer arithmetic algorithms, CRC Press (2nd edition) 2018
3. C. M. Bishop, Pattern Recognition and Machine Learning, 2016
Springer,Cambridge University Press (2nd edition)
4. M. Ercegovac and T. Lang, Digital arithmetic, Elsevier (1st edition) 2003
5. M. G. Arnold, Verilog digital computer design: algorithms into 1999
hardware, Prentice Hall (2nd edition)
6. H. Kaeslin, Digital integrated circuit design: from VLSI 2009
architectures to CMOS fabrication, Cambridge University Press (2nd
edition)
7. J. G. Delgado-Frias. and W. R. Moore, VLSI for neural networks 2013
and artificial intelligence, Plenum Press (1st edition)
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

NAME OF DEPARTMENT/CENTRE: Centre for Artificial Intelligence and Data Science

1. Subject Code: AID-583 Course Title: Data-driven Analytics for Smart Transportation Systems
2. Contact Hours: L: 3 T: 1 P: 0
3. Examination Duration (Hrs.): Theory: 3 Practical: 0
4. Relative Weightage: CWS: 20-35 PRS: 0 MTE: 20-30 ETE: 40-50 PRE: 0
5. Credits: 4 6. Semester: Both 7. Subject Area: PEC
8. Pre-requisite: Nil
9. Objective: To familiarize with the applications of data science in traffic and transportation engineering
and to demonstrate the applications of the data science in smart transportation planning

10. Details of the Course:

S.No. Contents Contact

hours
1. Data Science in Transportation 4
Overview and Practical Applications; Transportation Data Sources; Data
Collection; Data Preparation and Visualization.
2. Sensing and Data Mining for Smart Transportation Systems 8
Intelligent Transportation Systems, Incident Management Program, Efficient
Emergency Vehicle Movement (Pre-Emption), Crash Detection, Reporting, and
Clearance; Traffic Surveillance, Identification of Hotspots, Violation Detection;
Road Network Asset Management, Inventory of Potholes, other Deficiencies;
Adaptive Traffic Signal.
3. Data Analytics in Urban Transportation Planning 10
Basics of Urban Transportation Planning, Data Collection and Advanced Data
Sources, Household Surveys, Demand Modeling using WiFi/ Bluetooth/ Call Data
Record, Data Extraction and Analysis using APIs, Trip Distribution Modelling
Approaches, Route Choice Models, Choice Set Generation Methods, Genetic
Algorithms, Transportation Planning Example using Data-Driven Modeling and
Simulation.
4. Urban Mass Transit System 6
Basics of Urban Mass Transit System, Static and Dynamic GTFS, Real-Time
Transit, Travel Time Variability, Transit Reliability, Transit Planning using Smart-
Card Data, Real-Time Accessibility Analysis.
5. Applications in Environmental Impact of Transport System 6
IOT based Air pollution, Real-Time Air Pollution Monitoring and Data Analysis,
Placement of Mobile Sensors, Pollution Prediction using ML, Noise Data, Analysis
of Key Parameters, Development of Policy Framework.
6. Crash Data Analytics 8
Crash Data, Data Preparation, Model Estimation, Real-Time Data-Driven
Analysis; Emergency Vehicle Data, Crash Prone Stretches, Ambulance
Deployment; Near-misses/Traffic Conflict Data, Surrogate Approach, Proactive
Assessment and Safety Interventions.
Total 42
11. Suggested books
S.No. Name of Authors/Book/Publisher Year of
Publication/ Reprint
1. Fumitaka Kurauchi, Jan-Dirk Schmöcker “Public transport planning 2021
with smart card data” CRC Press
2. Juan de Dios Ortúzar, Luis G. Willumsen “Modelling Transport”, 2011
Wiley
3. Vukan R. Vuchic “Urban Transit: Operations, Planning, and 2005
Economics” Wiley
4. Constantinos Antoniou, Loukas Dimitriou, Francisco Pereira 2018
“Mobility Patterns, Big Data and Transport Analytics” Elsevier
5. Sara Moridpour, Alireza Toran Pour, Tayebeh Saghapour “Big Data 2019
Analytics in Traffic and Transportation Engineering: Emerging
Research and Opportunities” IGI Global
6. Khaled R. Ahmed, Aboul-Ella Hassanien “Deep Learning and Big 2021
Data for Intelligent Transportation” Springer
7. Davy Janssens, Ansar-Ul-Haque Yasar and Luk Knapen “Data 2013
Science and Simulation in Transportation Research” IGI Global