Home/Degree/CS3491
Back to Degree
CS3491Actively Used

Artificial Intelligence and Machine Learning

This course is part of the B.E. Computer Science Engineering curriculum under Anna University Regulation 2021. The knowledge from this course continues to be actively applied in professional software development.

Semester 4 (Second Year)
4 Credits
60 Lecture Hours

Course Overview

  • UniversityAnna University
  • Regulation2021
  • Semester4
  • Credits4
  • TypeCore
  • Units5

Course Objectives

1

To understand AI problem-solving techniques

2

To learn probabilistic reasoning and Bayesian networks

3

To understand supervised learning algorithms

4

To learn ensemble techniques and unsupervised learning

5

To implement neural networks and deep learning

Syllabus

Detailed unit-wise breakdown of the course curriculum as per Anna University Regulation 2021.

1

PROBLEM SOLVING

12 Hours
Introduction to AI and applicationsProblem solving agentsSearch algorithmsUninformed search strategiesHeuristic search strategiesLocal search and optimizationAdversarial searchConstraint satisfaction problems (CSP)
2

PROBABILISTIC REASONING

12 Hours
Acting under uncertaintyBayesian inferenceNaïve Bayes modelsProbabilistic reasoningBayesian networksExact inference in BNApproximate inference in BNCausal networks
3

SUPERVISED LEARNING

12 Hours
Introduction to machine learningLinear Regression – Least squaresSingle & multiple variablesBayesian linear regressionGradient descentLinear Classification – Discriminant functionLogistic regressionNaive Bayes classifierSupport vector machineDecision TreeRandom forests
4

ENSEMBLE TECHNIQUES AND UNSUPERVISED LEARNING

12 Hours
Combining multiple learnersModel combination schemesVotingEnsemble Learning – Bagging, Boosting, StackingUnsupervised learning – K-meansInstance-Based Learning – KNNGaussian mixture modelsExpectation maximization
5

NEURAL NETWORKS

12 Hours
PerceptronMultilayer perceptronActivation functionsNetwork training – Gradient descent optimizationStochastic gradient descentError backpropagationFrom shallow to deep networksVanishing gradient problemReLUHyperparameter tuningBatch normalizationRegularization and dropout

Course Outcomes

Upon completion of this course, students will be able to:

CO1

Apply AI search algorithms for problem solving

CO2

Use probabilistic reasoning for uncertainty handling

CO3

Implement supervised learning algorithms

CO4

Apply ensemble and unsupervised learning techniques

CO5

Build and train neural networks

Industry Application & Relevance

How the concepts learned in this course are applied in real-world software development projects across Banking, Healthcare, and Enterprise domains over 20+ years of experience.

Professional Application

AI integration, intelligent systems, predictive analytics

Textbooks & References

Textbooks

  • Stuart Russell, Peter Norvig, 'Artificial Intelligence: A Modern Approach', Pearson
  • Ethem Alpaydin, 'Introduction to Machine Learning', MIT Press

Reference Books

  • Tom Mitchell, 'Machine Learning', McGraw Hill
  • Christopher Bishop, 'Pattern Recognition and Machine Learning', Springer