Course Title: Artificial Intelligence and Machine Learning
Type of Course: Optional, Theory
Offered to: EEE
Pre-requisite Course(s): None
Introduction to Artificial Intelligence (AI): perception and intelligence, history; intelligent agent, algorithms in AI; ethical AI. Search and Optimization: graph search, uniform search, heuristic search, adversarial search, local search with constraint satisfaction. Logical Intelligence: logical agents, propositional logic, syntax, semantics, logical statement, first order logic. Introduction to Machine Learning: supervised, unsupervised, and reinforcement learning; components of the learning problem. Data mining and statistical pattern recognition. Learning models: linear classification and linear regression; extending linear models through nonlinear transforms, logistic regression, maximum likelihood, and gradient descent. Supervised learning: parametric/non-parametric algorithms; support vector machines; kernels. Unsupervised learning: clustering; dimensionality reduction; recommender systems. Deep learning and neural networks: multi-layer perceptron, backpropagation; convolutional networks; recurrent networks; attention mechanism and transformers. Best practices in machine learning: bias/variance theory; hyperparameter tuning. Case studies and applications.
To understand the fundamentals of AI and machine learning algorithms
To be able to implement AI based algorithms to solve real-life problems
To analyze various challenges in implementing machine learning and deep learning algorithms
To design machine learning and deep learning algorithms to solve real life applications
Computer programming and fundamental mathematics courses.
| CO No. | CO Statement | Corresponding PO(s)* | Domains and Taxonomy level(s)** | Delivery Method(s) and Activity(-ies) | Assessment Tool(s) |
|---|---|---|---|---|---|
| 1 | understand the fundamentals of AI and machine learning algorithms with real life applications | PO(a), PO(b) | C2 | Lectures, Discussions | Assignment, Class test, Final exam |
| 2 | solve real-life problems by designing suitable AI based algorithm | PO(d) | C4 | Lectures, Discussions | Assignment, Class test, Final exam |
| 3 | analyze real life challenges in implementing supervised and unsupervised learning algorithms | PO(b) | C4 | Lectures, Discussion, Demonstration | Assignment, Class test, Final exam |
| 4 | apply knowledge of regression analysis for effective recommendation | PO(d) | C3 | Lectures, Discussion, Demonstration | Assignment, Class test, Final exam |
| 5 | design deep learning models suitable for performing classification task | PO(c) | C6 | Lectures, Discussion, Demonstration | Assignment, Class test, Final exam |
| 6 | experience real life applications of ML and DL techniques | PO(b) | C4 | Lectures, Discussion, Demonstration | Assignment, Class test, Final exam |
Cognitive Domain Taxonomy Levels: C1 – Knowledge, C2 – Comprehension, C3 – Application, C4 – Analysis, C5 – Synthesis, C6 – Evaluation, Affective Domain Taxonomy Levels: A1: Receive; A2: Respond; A3: Value (demonstrate); A4: Organize; A5: Characterize; Psychomotor Domain Taxonomy Levels: P1: Perception; P2: Set; P3: Guided Response; P4: Mechanism; P5: Complex Overt Response; P6: Adaptation; P7: Organization
Program Outcomes (PO): PO(a) Engineering Knowledge, PO(b) Problem Analysis, PO(c) Design/development Solution, PO(d) Investigation,
PO(e) Modern tool usage, PO(f) The Engineer and Society, PO(g) Environment and sustainability, PO(h) Ethics, PO(i) Individual work and team work,
PO(j). Communication, PO(k) Project management and finance, PO(l) Life-long Learning
* For details of program outcome (PO) statements, please see the departmental website or course curriculum
| K1 | K2 | K3 | K4 | K5 | K6 | K7 | K8 | P1 | P2 | P3 | P4 | P5 | P6 | P7 | A1 | A2 | A3 | A4 | A5 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Week | Lectures | Topic |
|---|---|---|
| 1 | 1-3 | Introduction to AI and machine learning, history of AI. Intelligent agent: perception and intelligence, rational agent, reflex agent, problem solving agent. Ethical AI and biases. |
| 2 | 4-6 | Search and Optimization: graph search, tree search, uniform search strategies, breadth-first search, depth-first search, bidirectional search. Heuristic search: greedy search, A* search; |
| 3 | 7-9 | Local search with constraint satisfaction; adversarial search. Logical intelligence: logical agents, knowledge based agents. |
| 4 | 10-12 | Propositional logic, syntax, semantics, logical statement, truth table enumeration, first order logic. |
| 5 | 13-15 | Supervised, unsupervised and semi-supervised learning, reinforcement learning, components of the learning problem, relationship between in-sample and out-of-sample, K-nearest neighbour classifier. |
| 6 | 16-18 | Introduction with unsupervised learning, K-means clustering, hierarchical clustering, clustering evaluation. Dimensionality reduction, feature extraction, principle component analysis, feature selection: filtering and wrapper method. |
| 7 | 19-21 | Introduction to linear classification and linear regression, extending linear models through nonlinear transforms. |
| 8 | 22-24 | Parametric/non-parametric algorithms, support vector machines, introduction to kernels. |
| 9 | 25-27 | Introduction to logistic regression, maximum likelihood, gradient descent, recommender systems. |
| 10 | 28-30 | Introduction to data mining, statistical pattern recognition. |
| 11 | 31-33 | Multi-layer perceptron, backpropagation, convolutional neural networks. |
| 12 | 34-36 | Recurrent networks, attention mechanism, augmentation and transformers. |
| 13 | 37-39 | Bias/variance theory, hyperparameter tuning, segmentation architecture. |
| 14 | 40-42 | Case studies: application of learning algorithms to building smart robots (perception, control), computer vision, medical informatics, voice/audio, image database and other areas. |
Class participation will be judged by in-class evaluation; attendance will be recorded in every class.
Continuous assessment will be done in the form of quizzes, assignments, in-class evaluations.
Final Examination: A comprehensive term final examination will be held at the end of the Term following the guideline of academic Council.
Class Participation 10%
Continuous Assessment 20%
Final Examination 70%
Total 100%
Artificial Intelligence: A Modern Approach by Stuart Jonathan Russell and Peter Norvig.
Kernel Methods and Machine Learning by Sun Yan Kung
Deep Learning by Ian Goodfellow, Yoshua Bengio and Aaron Courville.
Artificial Intelligence: A New synthesis by Nils J. Nilsson.
Pattern Recognition and Machine Learning by Christopher M. Bishop.
Introduction to Machine Learning, Second Edition by Ethem Alpaydin
Online resources or supplementary materials will be shared with the class on a need basis
Besides going through relevant topics of the textbook, it is strongly advised that the students follow the class Lectures and discussions regularly for a thorough understanding of the topics.