Course Details

Subject {L-T-P / C} : EC6609 : Soft Computing { 3-0-0 / 3}

Subject Nature : Theory

Coordinator : Prof. Upendra Kumar Sahoo

Syllabus

Module 1: Fundamental Concepts: Introduction to Artificial Neural Networks (ANN). Learning Process: Error–correction learning, Hebbian learning, competitive learning, Boltzmann learning, the credit-assignment problem, supervised learning, and other learning techniques. [6 Hrs]

Module 2: Single neuron/ Perceptron networks: Training methodology, typical application to linearly separable problems. [3 Hrs]

Module 3: Multilayer Perceptron: Back propagation algorithm, virtues and limitation of BP algorithm, modifications to back-propagation. [10Hrs]

Module 4: Radial-basis function Networks – interpolation problem, Covers theorem, regularization networks, applications. Recurrent Networks. [6Hrs]

Module 5: Introduction to Fuzzy systems, Membership function, Fuzzy relational operation, fuzzy IF THEN rules, Sugeno and Mamdani type systems, Adaptive Neuro-Fuzzy systems, training methods Application of ANN and Fuzzy systems to non-stationary time series prediction pattern classification control communication engineering system identification. [12 Hrs]

Course Objectives

Course Outcomes

CO1: Able to carry out research and development of the neural network and fuzzy logic. <br />CO2: Become well aware on variety of soft computing techniques, their application in diverge fields. <br />CO3: Able to do in-depth analysis on technology variations in soft computing depending on the application and involvement of neural learning, classification. <br />CO4: Acquire improvise knowledge on multilayer perceptron, radial basis function, fuzzy logic. <br />CO5: Application of neural network and fuzzy logic for different applications like time series prediction, control.

Essential Reading

Supplementary Reading