Artificial Intelligence

Wester Zela
Session Three
[email protected]
www.ssyspe.org
Agenda

• Machine Learning
• Applications of Machine Learning
• Supervised Learning
• Unsupervised Learning
• Reinforcement Learning
• Steps to development an Machine Learning application
Machine Learning
• A Field of study that gives computers the ability to learn without
being explicitly programmed.
• A computer program is said to learn from experience E with respect
to some class of tasks T and performance measure P, if its
performance at tasks in T, as measured by P, improves with the
experience E (Tom M. Mitchell).
• Machine learning lies at the intersection of computer science,
engineering, and statistics and often appears in other disciplines.
• It can be applied to many fields from politics to Medicine.
Machine Learning
• Started with Alan Turing in 1950 … “The imitation game”, Can
machines think?. (Computing machinery and intelligence. Mind,59,
433-560).
• Loebner Prize (https://aisb.org.uk/events/aisb-convention-2020-
communication-and-conversation-non-members/)
• https://www.chatbots.org/ai_zone
• http://www.ssyspe.org/viewtopic.php?f=37&t=3896
• Today what is happening in your normal day ?…
• https://www.youtube.com/watch?v=MK31E4mSbXw
 Some Applications of Machine Learning
• Spam detection
• ChatBots
https://dialogflow.com/
https://wit.ai/
https://rasa.com/
• Biometric recognition
http://facex.io/
https://docs.opencv.org/3.0-beta/modules/face/doc/facerec/facerec_tutorial.html
https://github.com/davisking/dlib/blob/master/python_examples/face_recognition.py
https://cmusatyalab.github.io/openface/
http://openbiometrics.org/docs/index.html
• Stock Trading
Some Applications of Machine Learning
• Object recognition
http://opendetection.com/
• Advertising
https://www.youtube.com/watch?v=uiDMlFycNrw
• Medicine and Healthcare
https://www.youtube.com/watch?v=_Xcmh1LQB9I
• Retail and e-commerce
https://www.youtube.com/watch?v=F4chm96TYVE
• Games
https://gym.openai.com/
Some Applications of Machine Learning
• Internet of things
• Raspberry Pi
• Arduino Projects
https://store.arduino.cc/product/K000007
https://www.youtube.com/watch?v=5Z7z3Zduj6w
Supervised Learning
• Is the machine learning task of inferring a function from labeled
training data.
• We tell to the algorithm what to predict.
• Data
• Training Data
• Training set is a set of training examples to train machine learning algorithms
• Target Variable are known
• Learns by finding some relationship between the features and target variable
• Test set
• Data used to validate the algorithm
Supervised Learning
• Knowledge representation: The knowledge representation may be in
the form of a set of rules; it may be a probability distribution or an
example from the training set.
• Some Tasks in Supervised Learning:
• Clasification:predict what class an instance of data should fall into.
• Regression : is the prediction of a numeric value
Supervised Learning
Bird species clasification based on four features
Supervised Learning
• Some Algorithms in Supervised Learning:
• Decision trees
• Artificial Neural Network
• Deep Learning
• Genetic Algorithm
• Bayesian Networks
• Support Vector Machines (SVM)
• K-nearest neighbor
• Hidden Markov Models
Unsupervised Learning
• In unsupervised learning, there’s no label or target value given for the
data.
• The problem of Unsupervised Learning is try to find hidden structure
in unlabeled data.
• There is no error or reward signal to evaluate a potential solution.
• Some tasks:
• Clustering: A task where we group similar items together or dividing a large
data set into smaller data sets of some similarity
Unsupervised Learning - Clustering
Algorithms:
• K-Means Clustering
Supervised or Unsupervised Learning?
• The mesaure of amount of pollution in certain locations, under different weather conditions,
different time of days and different days of the week
• Predict the air pollution concentration at new location with known weather and known date
and time
Supervised or Unsupervised Learning?

Detection of communities in social networks

• We had to come up with an algorithm that could
divide the social network into clusters
• This algorithm should scale up for graphs
containing millions of nodes
• The quality measure should be shuch that it helps
analysis of desease spread in the network
Reinforcement Learning
• Concerned with how software agents should take actions in an
environment so as to maximize some notion of cumulative reward.
• Reinforcement learning differs from standard supervised learning in
that correct input/output pairs are never presented.
• There is a focus on on-line performance, which involves finding a
balance between exploration (of uncharted territory) and exploitation
(of current knowledge)
https://www.youtube.com/watch?v=VMp6pq6_QjI
https://gym.openai.com/
Reinforcement Learning
• Some Algorithms
• Temporal Difference Learning (TDL)
• Q-Learning
• Others….
Others Algorithms
• Evolutionary Algorithms
• Competitive Evolution
https://www.youtube.com/watch?v=LqUFJhkIqnI
• Cooperative Evolution
• Aplication to Go Game
Languages and Some tools
• C++
• Python: (check the book Machine Learning, Hands-On for
Developers andTechnical Professionals, Jason Bell, Willey)
• Hadoop
• Pyrobots: http://pyrorobotics.com/?page=Pyro
• Java
• Weka: http://www.cs.waikato.ac.nz/~ml/weka/
• https://www.tensorflow.org/?hl=es
• R, Visual R
Python for Machine Learning
Languages and Some tools
• Some tools and frameworks:
• Anaconda
• https://www.tensorflow.org/
• https://gym.openai.com/
• https://keras.io/
• https://github.com/deepmind/lab
• https://pytorch.org/
• .NET ( check the book: Machine Learning Projects for .NET
developers, Mathias Brandewinder, Apress)
• Matlab
• Hadoop
• SpringXD
Steps in development an Machine Learning
Application
• Start with some questions
• Collect the data
• Some data repositories:
• Ssyspe.org: http://www.ssyspe.org/viewforum.php?f=39
• Kaggle: https://www.kaggle.com/competitions
• Machine Learning Repository – UCI: http://archive.ics.uci.edu/ml/
• https://www.topcoder.com/
• http://datosabiertos.gob.pe/
• https://www.kaggle.com/datasets
• https://archive.ics.uci.edu/ml/datasets.html
• https://en.wikipedia.org/wiki/List_of_datasets_for_machine_learning_research
• https://dataverse.harvard.edu/
• http://reshare.ukdataservice.ac.uk/
Steps in development a Machine Learning
Application
• Prepare input data, If it is needed, clean data
• Check formats, length, missing fields, etc.
• Different Types of Data Format: Raw Text, Comma Separated Variables, JSON,
XML, Spreadsheets, Images, others.
• Analyze the Input data
• Find patterns, plotting data, etc.
• Normalize input data
• Data Scientists: Mathematics, Statisics, Programmer, Graphic Designer.
Steps in development a Machine Learning
Application
• Prepare for Data Processing and Data Storage (some are including
some AI libraries)
• https://aws.amazon.com/es/ec2/
• https://cloud.google.com
• https://azure.microsoft.com/es-es/
• Train the algorithms selected for the system
• Select and code the algorithm
• Use the 80% of the data for training
• In case of unsupervised learning, no training step because of we don’t have
target value.
Steps in development a Machine Learning
Application
• Test the algorithm
• Test the algorithms with 20% of data
• Select the algorithm with better results (i.e. % Correctly Classified Instances)
• In case of unsupervised learning, use other metrics to evaluate the success.
• Refine the algorithm
• Refine the algorithm (i.e. change some parameters, changes some input
variables, others)
• Move to production and monitor the results.
• Re-train in case of necesary.
Revision of Final Project
• Group 1:
• Group 2:
• Group 3:
• Group 4:
 Ranking in ssyspe.org: 13/01
• PTOi = 2*NPi *(Vistas/Total Vistas) + 1*NPAi/Total NP
Paper of Week
• https://towardsdatascience.com/machine-learning-101-supervised-
unsupervised-reinforcement-beyond-f18e722069bc

• Cluster analysis in retail segmentation for credit scoring, Sanja

Scitovski, and Natasa Sarlija, University of Osijek.

• Supervised Machine Learning: A Review of Classification

Techniques, S. B. Kotsiantis Department of Computer Science and
Technology, University of Peloponnese, Greece.
References
• Machine Learning (Chapter 1 and Chapter 2), Hands-On for Developers
andTechnical Professionals, Jason Bell, Willey.
• Machine Learning in Action, Peter Harrington, Manning Shelter Island
• Machine Learning Projects for .NET developers, Mathias Brandewinder, Apress