IITK Online PG Program

Module ID	Module Title	Credit	Description	Content
EE950	Data Analytics & Data Structures (DADS)	5	This course is a hands-on introduction to basic concepts in data analytics, data structures and visualization. The course provides the students with a comprehensive introduction to programming using Python and shell scripting, enabling them to work in a Linux environment, access remote servers and effectively debug their code. Additionally, the course aims to extensively cover data structures, including their implementation, manipulation, and analysis, while also teaching concepts such as file I/O formats, data readers, data visualization techniques like t-SNE, and the concept of Big-O notation. By the end of the course, students will have gained the necessary knowledge and tools to analyze data effectively using Python and navigate the Linux environment.	● Introduction and Preliminaries: Introduction to programming in Python Shell scripting Working in Linux Accessing remote servers Debugging ● Data reading: File I/O formats Data readers ● Visualization: Data Visualization t-SNE ● Data structures: Big-O notation Data structures
EE951	Introduction to Linear Algebra	5	This is an introductory linear algebra course that aims to provide students with a solid foundation in mathematical concepts and techniques relevant to machine learning. This course covers basic linear algebra topics such as vectors and matrices, singular value and other decompositions, solving systems of equations, linear independence, eigenvalue decomposition, and positive definite matrices.	●Vectors, vector operations, vector spaces, matrices, basic matrix operations, matrix multiplication ●Inner products, norms, linear functions ●Linear systems, LU and QR factorization ●Singular Value Decomposition, Spaces associated with a matrix ●Linear independence, Basis and Dimension ●Solving Ax = b, Determinant ●Eigenvalues, Eigenvalue decomposition ●Positive Definite Matrices ●Matrix calculus
EE952	Introduction to Machine Learning	5	This course aims at introducing the students to Machine Learning techniques used for various engineering applications. The lectures will focus on mathematical principles and there will be coding-based assignments for implementation, introducing students to tools such as sklearn and keras.	● Introduction to Preliminaries: Classification, Regression, Reinforcement Learning Evaluation Measures Basic Probability Theory ● Linear Model: Linear Regression Linear Classification ● Unsupervised Learning: Clustering Gaussian Mixture Model And visualization ● Supervised Learning: Regression Image Classification ● Time series Processing: Time series Analysis Dynamic Time warping ● ML at Scale: Parameter Tuning Model selection Validation and testing
EE953	Basics of Optimization	5	This is an introductory optimization course that seeks to introduce the various unconstrained optimization methods widely used in machine learning, particularly training of supervised models.	● Introduction and Preliminaries: Motivation Simple examples Local vs. global optimum Gradient of a function Numerical gradient ● Convexity: Convex Sets & functions Convex optimization problems Optimality Condition ● Gradient Descent: Narrative Optimization Gradient descent Line search Momentum ● Constrained Optimization: Constrained optimization Penalty methods ● Stochastic Gradient Descent: Stochastic gradient descent Implementation aspects
EE954	Deep Learning Fundamentals	5	The objective of the course is to provide students with a solid foundation in the principles, algorithms and techniques of deep learning. The course aims to enable students to understand and apply deep learning models, architectures, and training methodologies to solve complex problems in various domains such as computer vision, natural language processing, and data analytics.	● Introduction to Deep Learning Overview of neural networks and deep learning Historical development and key milestones Applications of deep learning in various domains ● Artificial Neural Networks Perceptron and multilayer perceptron Activation functions and feedforward propagation Backpropagation algorithm and gradient descent ● Optimization Algorithms Stochastic gradient descent (SGD) Adaptive optimization methods (e.g., Adam, RMSprop) Regularization techniques (e.g., dropout, L1/L2 regularization) ● Convolutional Neural Networks (CNNs) Motivation and architecture of CNNs Convolutional layers and pooling operations Training CNNs for image classification ● Recurrent Neural Networks (RNNs) Introduction to sequential data processing Architecture of RNNs and recurrent cells Training RNNs for sequence modeling tasks Long Short-Term Memory (LSTM) Networks ● Generative Models Introduction to generative modeling Variational Autoencoders (VAEs) Generative Adversarial Networks (GANs) ● Transfer Learning and Pretrained Models Introduction to transfer learning Using pre-trained models for new tasks
EE955	Probability and Statistics for Machine Learning	5	This course aims to provide fundamentals of probability theory and statistics required for machine learning. It’s designed for machine learning students and aims to provide them with a solid mathematical foundation in probability theory and statistics essential for ML applications. The course covers random variables, their distributions, statistics, and estimation.	● Introduction Introduction to Probability Random Variables ● Distributions Distribution (CDF) Probability Mass Function and Probability Density Function Conditional Probability and Independence The Law of Total Probability Variance and the Expected Value Covariance and Correlation ● Examples of Probability Distributions Discrete RVs e.g. Bernoulli, Binomial, Poisson Distribution Continuous RVs, e.g. Normal, Uniform, Gamma Multivariate Distribution ● Statistics Limit theorems, The Law of Large Numbers The Central Limit Theorem, Deviation Descriptive Deviation Bayesian Inference Estimation of RVs Maximum Likelihood Estimation Maximum Likelihood Estimation for Gaussian Distributions Confidence Intervals Hypothesis Testing and P-Values Chi-Square Test for Independence and Goodness of Fit
EE956	Machine Learning for Audio Processing	5	This course aims at introducing the students to machine learning (ML) techniques used for various audio processing applications. There will be spectral processing techniques for analysis and transformation of audio signals. The lectures will focus on mathematical principles, and there will be coding-based assignments for implementation.	● Introduction to speech and music: Speech and languages Music: Indian and western ● Digital signal processing: Digital signal processing basics Fourier Transforms Pitch and melody ● Machine Learning Review: Machine Learning basics Neural Networks (Dense, CNN, RNN, LSTMs) ● Audio Classification: Audio embeddings Radio Scene Classification ● Automatic Speech Recognition (ASR): Acoustic and Language models GMM-HMM based ASR DNN-HMM based ASR End-to-end deep ASR ● Music Information Retrieval: Music transcription Music tagging ● Audio Search: Embeddings and Hashing Search methods
EE957	Computer Vision	5	This course provides an introduction to the field of computer vision, focusing on the fundamental concepts, algorithms and applications. Students will learn about image processing, feature extraction, object detection, recognition and tracking. The course will also cover deep learning techniques for computer vision tasks. Through lectures and programming assignments, students will gain hands-on experience in developing computer vision applications.	● Introduction: Overview of computer vision Applications of computer vision Human visual system Basics of image formation ● Camera: Camera and Photography Camera control concepts Digital image creation: Quantization & Sampling ● Image manipulation: Spatial domain: Image enhancement Frequency domain: Image enhancement ● Edge detection: Canny edge Hough transform ● Segmentation: Thresholding Clustering based segmentation ● Image Compression: JPEG compression JPEG 2000 ● Image feature representation: Feature detector Feature representation Image features: LBP, Haar ● Convolutional neural networks: CNN and deep learning Feature representation learning CNN for object detection and recognition Applications and introduction to modern computer vision
EE958	Natural Language Processing	5	Natural language (NL) refers to the language spoken/written by humans. NL is the primary mode of communication for humans. With the growth of the world wide web, data in the form of text has grown exponentially. It calls for the development of algorithms and techniques for processing natural language for the automation and development of intelligent machines. This course will primarily focus on understanding and developing techniques, statistical learning algorithms, and models for processing language. We will have a statistical approach towards natural language processing, wherein we will learn how one could develop natural language understanding models from statistical regularities in large corpora of natural language texts while leveraging linguistic theories.	● Introduction: Introduction Why is NLP hard? Linguistics fundamentals ● Language Models, tagging, and parsing: Language Models: n-grams, smoothing, class-based, brown clustering Sequence Labeling: HMM, MaxEnt, CRFs, Part of Speech tagging Parsing: CFG, Lexicalized CFG, PCFGs, Dependency parsing ● Applications: Named Entity Recognition Coreference Resolution Text classification Toolkits e.g. Spacy ● Advanced Topics: Distributional Semantics, vector space models Distributed Representations: Neural Networks, Backpropagation, Softmax Word Vectors: Word2Vec, GloVE, FastText, ELMo Deep Models: RNNs, LSTMs, Attention, CNNs Sequence to Sequence models: machine translation Transformers: BERT, transfer learning Graph Neural Networks: GCN, architecture and applications
EE959	ML with Large Datasets	5	This is an introductory optimization course that seeks to introduce various unconstrained optimization methods widely used in machine learning, particularly for training supervised models.	● Introduction and Spark Preliminaries: Distributed Computing, Databricks Visualization, Dimensionality Reduction Distributed Linear Regression ● Basic Algorithms: Kernel approximations Logistic regression, hashing Distributed trees ● Deep Learning: Deep learning, automatic differentiation Large Scale Optimization Optimization for DL Hyperparameter tuning ● Distributed Learning: Parallel distributed DL Federated Learning ● Advanced Topics: Neural architecture search Model compression
EE960	AI in IoT	5	The objective of the course is to equip students with the knowledge and skills to effectively apply artificial intelligence techniques in the context of Internet of Things (IoT) systems. By covering the basics of IoT communication and security aspects alongside AI applications, the course aims to enable students to design, develop, and deploy intelligent IoT solutions that leverage the power of AI algorithms and models.	● Introduction to IoT: New Trends and Applications IoT Architecture Middleware Fog Computing Sensors and Actuators ● IoT Communications and Sensor Networks: NFC, RFID Bluetooth, Zigbee, WiFi MQTT, HTTP Network Topologies Challenges, Routing, and Optimization ● IoT Security: Device Security Communication Security Digital Forensics ● AI in IoT: Smart Cities Healthcare Agriculture Manufacturing
EE961	AI in Healthcare	5	This course explores the applications of artificial intelligence (AI) in the healthcare domain. Students will learn about the fundamental concepts of AI, machine learning, and deep learning and how they are applied to various healthcare tasks. The course will cover topics such as medical image analysis, clinical decision systems, electronic health records, and personalized medicine.	● Introduction: Overview of AI and its impact on healthcare Ethical considerations in AI-driven healthcare Challenges and opportunities in AI adoption Machine learning fundamentals ● Visual data analysis in medical domain: Feature engineering and feature selection for medical domain Introduction to medical imaging modalities (e.g., X-ray, MRI) Image data formats in medical domain Image segmentation and feature extraction Deep learning for medical image analysis Instruments and sensor analysis in medical domain ● ML and rule-based disease diagnosis: Role of AI in clinical decision making Rule-based systems and expert systems Machine learning models for diagnosis and prognosis Explainability and interpretability in clinical decision support ● Healthcare records processing and robotics in healthcare: Overview of electronic health records Data mining techniques for EHR analysis Predictive modeling using EHR data AI approaches in drug discovery and development Virtual screening and molecular docking Genomic data analysis and personalized medicine Robotics applications in surgery and rehabilitation Surgical planning and assistance systems Human-robot interaction in healthcare settings
EE962	Industrial AI and Automation/ AI in Industry and Automation	5	This course explores the applications of artificial intelligence (AI) in industrial settings and automation processes. Students will learn about the use of AI techniques such as machine learning, robotics, sensory data, and image processing in various industries. The course will cover topics such as smart manufacturing, predictive maintenance, supply chain optimization, and intelligent automation. Through lectures, case studies, and hands-on assignments, students will gain an understanding of how AI is transforming industries and enabling efficient and intelligent automation.	● Introduction to AI in Industry and Automation: Overview of AI and its impact on industries Role of automation in industrial processes Challenges and opportunities in adopting AI in industry ● Robotics and Automation: Introduction to Industrial Robotics Robot kinematics and dynamics Robot control systems and programming Collaborative robots and human-robot interaction ● Industry 4.0: Concepts of smart manufacturing and Industry 4.0 AI-enabled quality control and defect detection Predictive maintenance and condition monitoring Digital twin technology and virtual commissioning ● Future Trends and Emerging Applications: Optimization of supply chain processes Demand forecasting using AI techniques Route optimization and fleet management Warehouse automation and inventory management Cognitive automation and decision support systems Workflow automation and business process optimization Fraud detection and risk assessment using AI Algorithmic trading and portfolio management Chatbots and virtual assistants in banking AI in agriculture and food production AI in transportation and autonomous vehicles
EE963	Reinforcement Learning	5	In this course we will explore how an agent (via interactions with the environment) can learn by trial and error. This is quite different from supervised machine learning and comes close to how humans learn by interactions. Reinforcement Learning (RL) deals with problems that require sequential decision making. This course will explore foundations of reinforcement learning. We will study different algorithms for RL and later in the course we will explore how functional approximation in RL algorithms could be done using neural networks giving rise to deep reinforcement learning.	● Introduction RL task formulation Action space, state space, environment ● Dynamic Programming Tabular based solution Dynamic Programming Monte Carlo Temporal Difference ● Functional Approximation and Deep RL Value based Deep Reinforcement Learning: Functional Approximation in RL, NFQ (Neural Fitted Q Iteration), DQN (Deep Q-Network), Double DQN, Dueling DDQN, PER (Prioritized Experience Replay) Policy Based and Value Based Algorithms: REINFORCE, Vanilla Policy Gradient (VPG), A3C (Asynchronous Advantage Actor Critic), Generalized Advantage Estimation (GAE), Advantage Actor-Critic (A2C), SARSA Advanced Actor Critic: DDPG (Deep Deterministic Policy Gradient), TD3 (Time Delayed DDPG), SAC (Soft Actor Critic), PPO (Proximal Policy Optimization) ● Advanced Topics Model-based RL Imitation Learning Meta-Learning Multi-agent Learning POMDP
EE964	Project	5	The goal of this module is to have the students do an industry-relevant project in a topic related to machine learning. A module may have one or more instructors, who will decide the topics of the projects in consultation with the enrolled instructors. The project will typically involve design and development of a novel ML model or algorithm. A specific project may be split across at most two modules.	● Sentiment analysis ● Advanced image classification ● Fraud Detection ● Recommendation Systems ● Spam Email Classification ● Disease Diagnosis ● Stock Price Prediction ● Object Detection ● Facial Emotion Recognition ● Natural Language Generation
EE965	Unsupervised Learning	5	The objective of this course is to introduce the students to unsupervised machine learning techniques used for various engineering applications. The students will gain the skills to extract valuable insights from datasets lacking a specified target or labeled variable. The lectures will focus on underlying mathematics principles as well as application problems in various domains. The students will also be introduced to unsupervised learning libraries such as sklearn.	● Introduction and K-means Clustering ● Hierarchical and Spectral Clustering ● Dimension Reduction - Linear and Non-linear ● Matrix Factorization, NMF optimization ● Graphical Models, Bayesian Networks, Markov Random Fields ● Mixture Models and EM ● Approximate Inference
EE966	AIML Projects with real-world datasets	5	As part of the course, students will participate and successfully complete several PYTHON-based projects and case studies on key AI/ML techniques such as Linear Regression, Logistic Regression, Support Vector Machines, Naïve Bayes, and more.	● Introduction to PYTHON, ML Packages, Data compression, Principal Component Analysis (PCA): Introduction to PYTHON libraries SCIKIT, PANDAS, NUMPY, Introduction to PCA algorithm, PCA via SVD Project 0: SCIKIT, PANDAS, NUMPY and other ML modules in PYTHON Project 1: PCA-Based clustering for IRIS dataset ● Linear Regression: Regression applications, Problem formulation and solution Project 2: IRIS Dataset Regression using PYTHON Project 3: Boston Housing Price Analysis using PYTHON-Based Regression Project 4: California Housing Price Analysis using PYTHON-Based Regression ● Logistic Regression: Logistic function, Likelihood maximization, Online learning for parameter estimation Project 5: SCIKIT Package for Logistic Regression using Purchase/Shopping Data Project 6: Logistic Regression application for Wine quality dataset ● Support Vector Machines: SVM applications, Maximum margin classifier, Kernel SVM Project 7: Breast Cancer Dataset Analysis using SVC Project 8: IRIS Dataset Classification using PYTHON-Based SVC ● Naïve Bayes: Discrete feature vectors, Naïve Bayes assumption, Posterior probabilities, Laplacian smoothing Project 9: Naïve Bayes Clustering of Purchase Dataset using SCIKIT library Project 10: Wine quality dataset classification using Naïve Bayes ● Linear Discriminant Analysis: Multivariate Gaussian modeling, Likelihood Ratio test, Discriminant function Project 11: Discriminant Based Data Classification using IRIS Dataset ● Decision Tree Classifiers (DTC): DTC structure, choice of best attribute, Entropy, Mutual Information or Information Gain Project 12: Building a DTC using the Purchase Logistic Dataset Project 13: Building a DTC for IRIS Dataset using PYTHON ● K-Means and Probabilistic Clustering: Unsupervised learning, K-Means procedure, EM Algorithm, Soft clustering Project 14: Clustering Analysis using PYTHON
EE967	Deep Learning and Neural Networks (DLNN) Projects with real-world datasets	5	As part of the course, students will participate and successfully complete several PYTHON (TensorFlow and Keras)-based projects and case studies on key Deep Learning (DL)/ Neural Network (NN) techniques such as Neurons, Activation Functions, Deep Neural Networks and Convolutional Networks in significant detail. These projects will be based on practical real-world datasets such as Boston Housing Price, California Housing Price, Mobile Phone Dataset, Fashion Dataset, Handwritten Digit Classification Dataset, IMDB movie rating and others. Another important aspect of the program is to study Structure of Neural Units, Properties of Neurons, Back Propagation, Convolution, Pooling and Flattening operations etc. Students will also develop the skills to effectively use integrated development environments (IDEs) in PYTHON and advanced packages such as TensorFlow and Keras for tackling more extensive DL/ NN projects in the future using PYTHON-Based Neural Networks.	● Introduction to PYTHON, DL Packages and Basics of Neural Networks: Introduction to PYTHON libraries, TensorFlow, Keras, Introduction to Neural Networks, Structure of Neural Nets, Properties of Neurons and Activation ● Projects on Neural Networks for Boston and California Housing Price Datasets: Project 1: Boston Housing Price Analysis using Neural Network Project 2: California Housing Price Analysis ● Project on Mobile Phone Price Analysis: Using Neural Networks to predict price range based on parameters such as clock speed, dual sim, carrier, 4G, memory, cores, pixel height, pixel width etc. ● Deep Learning, Multi-layer Neural Networks: Architecture of Deep Neural Networks, Mathematical Analysis of Back Propagation, Algorithm for Back Propagation with arbitrary number of layers ● Deep Learning Project for Fashion Classification: Using MNIST Fashion Dataset with 60,000 training and 10,000 testing images; each 28x28 grayscale, classified into 10 categories ● Deep Learning Project for Digit Classification: Using the MNIST dataset of handwritten digits to benchmark classification algorithms using deep learning models ● Convolutional Neural Networks (CNNs): Introduction to CNNs, Basic structure, Activation Maps, Pooling and Flattening operations ● Projects using CIFAR and IMDB Datasets: CIFAR-10 Dataset with 60,000 32x32 color images in 10 classes IMDB Movie Rating Classification using Deep Learning

Artificial Intelligence and Machine Learning (AIML)