Scribd
Upload
75 views2 pages

Machine Learning for Engineering and science applications - - Unit 14 - Week 11

The document outlines the Week 11 assignment for the NPTEL course 'Machine Learning for Engineering and Science Applications,' including questions related to generative adversarial networks, MAP estimation, and Bayesian regression. It provides details on submission deadlines and the ability to submit answers multiple times before the due date. Additionally, it includes links to course materials and resources for further learning.

Uploaded by

pakhil
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
75 views2 pages

Machine Learning for Engineering and science applications - - Unit 14 - Week 11

The document outlines the Week 11 assignment for the NPTEL course 'Machine Learning for Engineering and Science Applications,' including questions related to generative adversarial networks, MAP estimation, and Bayesian regression. It provides details on submission deadlines and the ability to submit answers multiple times before the due date. Additionally, it includes links to course materials and resources for further learning.

Uploaded by

pakhil
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

X

[email protected]

(https://swayam.gov.in)

(https://swayam.gov.in/nc_details/NPTEL)

NPTEL (https://swayam.gov.in/explorer?ncCode=NPTEL) » Machine Learning for Engineering and science applications (course)


Click to register for
Certification exam

Week 11 : Assignment 11
(https://examform.nptel.ac.in/2025_01/exam_form/dashboard)

If already registered, Your last recorded submission was on 2025-04-07, 10:48 IST Due date: 2025-04-09, 23:59 IST.

click to check your 1) In the generative adversarial network (GAN) framework, what are the two neural networks involved? 1 point
payment status
Encoder and decoder
Teacher and student
Generator and discriminator
Course outline
Classifier and regressor

About NPTEL ()
2) When a Gaussian prior is used in MAP estimation, what type of regularization does it correspond to? 1 point

How does an L1 regularization


NPTEL online
L2 regularization
course work ? ()
Elastic net regularization
MATLAB () Dropout regularization

Week 1 () 3) What is the input to the generator network in a GAN? 1 point

Week 2 () Training data


Random noise vector
Week 3 () Output from the discriminator
Encoded representation of the data
Week 4 ()

4) What is the ”bottleneck layer” in an autoencoder? 1 point


Week 5 ()
The layer with the largest number of neurons
Week 6 ()
The layer where dimensionality reduction occurs
The layer with the slowest computation time
Week 7 ()
The final output layer
Week 8 ()
5) What is a key advantage of using Variational Autoencoders over traditional autoencoders? 1 point
Week 9 ()
VAEs are computationally less expensive

Week 10 () VAEs provide more structured and meaningful latent representations


VAEs automatically classify the input data
Week 11 () VAEs require less training data

MLE, MAP and


6) What advantage might an ensemble of classifiers have over a single classifier for cardiac disease diagnosis? 1 point
Bayesian Regression
(unit?
Ensemble methods are always faster computationally
unit=24&lesson=113)
They can capture different aspects of the data and reduce prediction variance
Introduction to
They require fewer features to be extracted
Generative model
(unit? They eliminate the need for ground truth labels
unit=24&lesson=114)
7) Why might preprocessing to extract the region of interest be beneficial in the cardiac MRI analysis compared to 1 point
Generative Adversarial
Networks (GAN) (unit? using the full image directly?
unit=24&lesson=115)
It reduces noise and helps the CNN focus on relevant structures
Variational Auto-
It’s required for all CNN applications
encoders (VAE) (unit?
unit=24&lesson=116) It makes the data more colorful and visually appealing
Applications: Cardiac It reduces the number of classes to predict
MRI - Segmentation &
Diagnosis (unit? 8) What function does a generative model learn that a discriminative model doesn’t? 1 point
unit=24&lesson=117)

Applications: Cardiac P (Y |X) - posterior probability


MRI Analysis -
Tensorflow code P (X, Y ) - joint probability
walkthrough (unit?
unit=24&lesson=118) P (X) - data distribution

Week 11 Feedback
P (Y ) - class distribution
form : Machine
Learning for
Engineering and 9) What is the key difference between Bayesian regression and MAP estimation? 1 point
Science Applications
(unit?
Bayesian regression uses a different cost function
unit=24&lesson=173) MAP only uses normal distributions

Lecture Material : 11 Bayesian regression provides a predictive distribution rather than a point estimate
(unit? MAP doesn’t require gradient descent
unit=24&lesson=238)

Quiz: Week 11 : 10) What is the dice coefficient used for in medical image segmentation? 1 point
Assignment 11
(assessment?
As a regularization term
name=262) To measure overlap between predicted segmentation and ground truth
To detect edges in the images
Week 12 ()
To normalize pixel values
DOWNLOAD You may submit any number of times before the due date. The final submission will be considered for grading.
VIDEOS ()
Submit Answers

Text Transcripts ()

Book ()

Problem Solving
Session - Jan 2025
()

You might also like