# 2K6 ME 805(B) / 2K6 CS 805(C) : NEURAL NETWORKS & FUZZY LOGIC

**Module I (13 hours)**

Introduction to artificial neural networks – biological neurons – Mc Culloch and Pitts modals of neuron – types of activation function – network architectures – knowledge representation – learning process – error-correction learning – supervised learning – unsupervised learning – single unit mappings and the perceptron – perceptron convergencetheorem (with out proof) – method of steepest descent – least mean square algorithms – adaline/medaline units – multilayer perceptrons – derivation of the back-propagation algorithm**Module II (13 hours)**

Radial basis and recurrent neural networks – RBF network structure – covers theorem and the separability of patterns – RBF learning strategies – K-means and LMS algorithms – comparison of RBF and MLP networks – recurrent networks – Hopfield networks – energy function – spurious states – error performance – simulated annealing – the Boltzman machine – Boltzman learning rule – the mean field theory machine – MFT learning algorithm – applications of neural network – the XOR problem – traveling salesman problem – image compression using MLPs – character retrieval using Hopfield networks**Module III (13 hours)**

Fuzzy logic – fuzzy sets – properties – operations on fuzzy sets – fuzzy relations – operations on fuzzy relations – the extension principle – fuzzy measures – membership functions – fuzzification and defuzzification methods – fuzzy controllers – Mamdani and Sugeno types – design parameters – choice of membership functions – fuzzification and defuzzification methods – applications**Module IV (13 hours)**

Introduction to genetic algorithm and hybrid systems – genetic algorithms – natural evolution – properties – classification – GA features – coding – selection – reproduction – cross over and mutation operators basic GA and structure Introduction to Hybrid systems – concept of neuro-fuzzy and neuro-genetic system

**Reference books**

1 Simon Haykins, “Neural Network a – Comprehensive Foundation”, Macmillan College, Proc, Con, Inc

2 Zurada J.M., “Introduction to Artificial Neural Systems, Jaico publishers

3 Driankov D., Hellendoorn H. & Reinfrank M., “An Introduction to Fuzzy Control”, Norosa Publishing House

4 Ross T.J., “Fuzzy Logic with Engineering Applications”, McGraw Hill

5 Bart Kosko. “Neural Network and Fuzzy Systems”, Prentice Hall, Inc., Englewood Cliffs

6 Goldberg D.E., “Genetic Algorithms in Search Optimisation and Machine Learning”, Addison Wesley

7 Suran Goonatilake & Sukhdev Khebbal (Eds.), “Intelligent Hybrid Systems”, John Wiley