Course Details

Subject {L-T-P / C} : EC3604 : Soft Computing { 3-0-0 / 3}

Subject Nature : Theory

Coordinator : Prof. Samit Ari

Syllabus

Fundamental Concepts: Introduction to Artificial Neural Networks (ANN), Models of a Neuron, Network structure Learning Process: Error–correction learning, Hebbian learning, competitive learning, Boltzmann learning, the credit-assignment problem, supervised learning, and other learning techniques. Single neuron/ Perceptron networks: Training methodology, typical application to linearly separable problems. Multilayer Perceptron: Back propagation algorithm, virtues and limitation of BP algorithm, modifications to back-propagation. Radial-basis function Networks – interpolation problem, Cover’s theorem, Regularization networks, applications.
Introduction to Fuzzy systems, Membership function, Fuzzy relational operation, Fuzzy IF THEN rules, Fuzzy Reasoning, Sugeno and Mamdani type systems, Adaptive Neuro-Fuzzy sytems, training methods, Application of ANN and Fuzzy systems to non-stationary time series prediction pattern classification.

Course Objectives

Course Outcomes

1. Understand the fundamental concepts of Artificial Neural Networks (ANN) and the various learning techniques used in the field. <br />2. Acquire the knowledge and skills to train and apply Single Neuron/Perceptron networks to linearly separable problems. <br />3. Develop proficiency in the Back Propagation algorithm for training Multilayer Perceptron networks and understand its limitations. <br />4. Gain insight into the concept of Radial-basis function Networks, their applications, and the use of regularization networks to overcome interpolation problems. <br />5. Learn about Fuzzy Systems, their components, and their applications in a wide range of fields, including non-stationary time series prediction, pattern classification, control, communication engineering, and system identification.

Essential Reading

Supplementary Reading