Neural Networks and Fuzzy Logic

Introduction to neural networks -Motivation from Biological neuron, McCulloch Pitts Neuron, -Single layer perceptrons, Multi Layer Perceptrons (MLPs), Linearly Separable Boolean functions, Representation Power of MLPs, Activation functions - Sigmoid, Tanh, ReLU, Softmax. Loss functions, Training MLPs with backpropagation. Practical issues neural network training - The Problem of Overfitting, Vanishing and exploding gradient problems.

Overview of crisp sets, Fuzzy Sets and Uncertainty: Uncertainty and information, fuzzy sets and membership functions, chance verses fuzziness, properties of fuzzy sets, fuzzy set operations. Fuzzy Relations: Cardinality, operations, properties, fuzzy cartesian product and composition, fuzzy tolerance and equivalence relations, forms of composition operation.

Features and Various forms of membership functions, fuzzification, defuzzification to crisp sets and scalars. Development of membership functions- Membership value assignments: intuition, inference, rank ordering, neural networks, inductive reasoning.

Introduction, Mamdani Fuzzy models, Sugeno models, Tsukamatto fuzzy models Other considerations: input space partitioning - grid partition, Re partition, tree partition, Scatter partition. Fuzzy modeling.

Neuro- Fuzzy modelling:- Introduction, ANFIS Architecture, hybrid learning algorithm, learning methods with ANFIS and RBFN, ANFIS as a universal approximator. Generalised ANFIS - Introduction, multi IO systems, architectural comparison, Neuron functions for adaptive networks, Analysis of adaptive learning capability.