Syllabus

State-based model: Modelling a search problem; formulating state spaces and actions; Backtracking search and its special cases such as depth-first search; breadth-first search and depth-first search with iterative deepening; coding demonstration of backtracking search with details on complexity analysis; Dynamic programming coding demonstration of backtracking search with details on complexity analysis; Uniform cost search coding demonstration of backtracking search with details on complexity analysis;  A* search coding demonstration of backtracking search with details on complexity; Algorithms for learning costs of a search problem given a state space. 

Markov Decision Process (MDP) and Reinforcement learning: Modelling a search problem with stochasticity; Concepts of chance nodes and transition probability; Detailing on concepts such as policy; utility and value with formal definitions and concrete examples; Detailed discussion on policy evaluation; Extending the idea of policy evaluation to value iteration; Bellman’s equation; Transition from MDP to Reinforcement Learning (RL); Mone Carlo Approach to RL;  Model based and model-free Monte  Carlo; SARSA algorithm for RL; Q-Learning; Brief insight into Deep Reinforcement Learning; A practical demonstration of RL.

Adversarial Games: Modelling two-player zero-sum games; Concepts of expectimax and minimax; Evaluation of Games; Coding example with a simple game; Inequalities related to the Minimax policy and their interpretation; Modelling games with a component of randomness; The concept of Expectiminimax; Alpha-Beta pruning for game trees; Temporal difference learning with example; Game evaluation. 

Constraint satisfaction problem: Factor graph with example; Dynamic ordering; Arc consistency; Beam search; Local search and Conditional Independence; Variable elimination. 

Logic: Relating Logic with Natural Language; Syntax versus semantics; Propositional logic; Syntax of propositional logic; Interpretation function; Contingency Contradiction and entailment; Tell operation; Ask operation; Satisfiability; Model checking; Inference framework; Desiderata for inference rules; Resolution and Soundness of Resolution; Limitation of Propositional Logic; First-order logic.

Text Books

1. Stuart Russell and Peter Norvig, Artificial Intelligence: A Modern Approach, 4th US ed. 

2. Stanford Lecture Series: Artificial Intelligence: Principles and Techniques. 

3. Trevor Hastie, Robert Tibshirani, Jerome Friedman, The Elements of Statistical Learning

3. Ian Goodfellow, Yoshua Bengio and Aaron Courville, Deep Learning. 

Download Links (Require Access Request):  

• Class Slides    

• Python Codes       

Syllabus

Introduction: Biological Neuron, Idea of computational units, McCulloch–Pitts unit and Thresholding logic, Linear Perceptron, Perceptron Learning Algorithm, Linear separability. Convergence theorem for Perceptron Learning Algorithm. 

Feedforward Networks: Multilayer Perceptron, Gradient Descent, Backpropagation, Empirical Risk Minimization, regularization.

Deep Neural Networks: Difficulty of training deep neural networks, Greedy layerwise training. 

Better Training of Neural Networks: Newer optimization methods for neural networks (Adagrad, adadelta, rmsprop, adam, NAG), second order methods for training, Saddle point problem in neural networks, Regularization methods (dropout, drop connect, batch normalization). 

Convolutional Neural Networks: Architectures, convolution/pooling layers, LeNet, AlexNet.

Recurrent Neural Networks: Backpropagation through time, Long Short Term Memory, Gated Recurrent Units, Bidirectional LSTMs, Bidirectional RNNs. 

Generative models: Restrictive Boltzmann Machines (RBMs), Introduction to MCMC and Gibbs Sampling, gradient computations in RBMs, Deep Boltzmann Machines.

Deep Unsupervised Learning and Recent Trends: Autoencoders (standard, sparse, denoising, contractive, etc), VariationalAutoencoders, Adversarial Generative Adversarial Networks, Autoencoder and DBM, Multi-task Deep Learning, Multi-view Deep Learning. 

Applications of Deep Learning to Computer Vision: Image segmentation, object detection, automatic image captioning, Image generation with Generative adversarial networks, video-to-text with LSTM models. Attention models for computer vision tasks. 

 Applications of Deep Learning to NLP: Introduction to NLP and Vector Space Model of Semantics Word Vector Representations: Continuous Skip-Gram Model, Continuous Bag-of-Words model (CBOW), Glove, Evaluations and Applications in word similarity, analogy reasoning.

Text Books

1. Ian Goodfellow and YoshuaBengio and Aaron Courville, Deep Learning, MIT Press, 2016. 

2. Bishop, C. ,M., Pattern Recognition and Machine Learning, Springer, 2006. 

3. Raúl Rojas, Neural Networks: A Systematic Introduction, Springer, 1996. 

Download Links (Require Access Request):  

• Class Slides    

• Python Codes       

Syllabus

Introduction to Data Analysis & Visualization: 

Data Analytics, Role of Data Analytics, Components of Data Analytics, Types of Data Analytics, Data Analytics Techniques, Advantages of Data Analytics, Applications of Data Analytics, Data Analytics Tools. Data Visualization, Importance of Data Visualization, Categories of Data Visualization, Types of Data Visualization Techniques, Advantages and Disadvantages of Data Visualization, Difference Between Data Visualization and Data Analytics, Data Visualization Techniques, Data Visualization Tools, Benefits of Data Visualization Tools, Common features of Data Visualization Tool, Data Visualization Libraries, Data Visualization examples, Concerns Regarding Data Visualization.

Data Analysis and Visualization Using Python: 

Steps from data to knowledge, Data analysis and processing, Python libraries in data analysis, How to choose the right Python libraries Data Visualization (Line Plot, Bar Plot, Pie Chart, Box Plot, Histogram Plot, Scatter Plot):- Direct Plotting, Seaborn Plotting System, Matplotlib Plot Case Studies:- Data Gathering, Data Analysis, Data Visualization, Findings Advanced Visualization:- Simulation, Animation Machine Learning Models with scikit-learn:- Logistic Regression, Random Forest, Naive Bayes Classifier, Support Vector Machine, Gradient Boosting, Decision Tree,  KNN.

Visualization for deep learning: 

An Interactive overview of model analysis, Why do we want to visualize models?, What can we visualize? When is visualization most relevant? How can we visualize models? Tools and frameworks for model visualization Understanding, Visualizations, and Explanation of Deep Neural Networks, Deep Learning Visualization Methods, Visualizing a Neural Network using Keras Library, Visualize Deep Learning Models using Visualkeras, Why Model Visualization is Important? What are the benefits of Model Visualization?

Visualization Toolkit (VTK) : 

Introduction to Visualization Toolkit (VTK) for 3D computer graphics, Image processing and visualization, Visualization Pipeline, Isosurfaces, Volume rendering, Vector field visualization, And biological and medical data applications.

Data Visualization BI Tool: 

Introduction to Data Visualization with Tableau, Exploring Data Visualization with Tableau, What is Data Visualization, Exporting Data and Working with Tableau

Building Data Visualization BI Project With Tableau: 

BI Reporting Understanding, Report and Dashboard Template Document, Tableau Design and Development Database Source Connection 

Text Books

1. 1. Ossama Embark, Data Analysis and Visualization Using Python: Analyze Data to Create Visualizations for BI Systems, Apress, 2018.

   2. Kieran Healy, Data visualization: A practical introduction, Princeton University Press, 2019.

   3. Tristan Guillevin, Getting Started with Tableau, Packet Publishing, 2019.

   4. Hansen, C.D., and Johnson, C.R., Visualization Handbook, Academic Press, 2004.

Download Links (Require Access Request):  

• Class Slides    

• Python Codes