MSc (Mathematics and Computing) Programme:Artificial Neural Networks

Semester I

Real Analysis – I
Linear Algebra
Complex Analysis
Fundamentals of Computer Science and C Programming
Discrete Mathematical Structure
Differential Equations


Semester II

Real Analysis –II
Advanced Abstract Algebra
Computer Oriented Numerical Methods
Data Structures
Data Based Management Systems
Operating Systems


Semester III

Topology
Computer Based Optimization Techniques
Computer Networks
Mechanics
Seminar


Semester IV

Functional Analysis
Dissertation


Artificial Neural Networks

Introduction: Biological Analogy, Architecture classification, Neural Models, Learning Paradigm and Rule, single unit mapping and the perception.

Concepts in ANN: Feed forward networks – Review of optimization methods, back propagation, variation on back propagation, FFANN mapping capability, Mathematical properties of FFANN’s Generalization, Bios and variance Dilemma, Radial Basis Function networks.

Recurrent Networks: Symmetric Hopfield networks and associative memory, Boltzmann machine, Adaptive Resonance Networks

Other Networks: PCA, SOM, LVQ, Hopfield Networks, Associative Memories, RBF Networks, Applications of Artificial Neural Networks to Function Approximation, Regression, Classification, Blind Source Separation, Time Series and Forecasting.

Laboratory Work: The lab work will be based on the implementations of different neural networks strategies using C/C++/LISP/PROLOG (or on MATLAB/MATHEMATICA) on various case studies.