Curriculum

Unit-I

Review of Linear algebra: Linear combination of vectors, linearly dependent and independent set of vectors, Vector space, subspace, basis, rank, Eigen vectors, orthogonal vectors, inner product, outer product.(No questions will appear in the end exam from these topics)

Basics of Artificial Neural Networks: Trends in computing, Pattern and Data, Pattern recognition tasks. Basic methods of pattern recognition, Basics of Artificial Neural Networks, Biological Neural Network, Models of neuron: McCulloch-Pitts Model, Perceptron, Adaline, topology, Supervised and unsupervised learning, Basic learning laws, Realization of logic functions using MP neuron 7 Hours

Unit-II

Functional units of ANN & Single layer perceptron: Basic ANN Models (architectures) for Pattern recognition task, Pattern recognition tasks by i) Feed-forward ii) Feed-back iii) competitive learning Neural networks. Feed-forward neural network: Linear associative network, Analysis of pattern classification networks, Linear separability, Perceptron convergence theorem. 7 Hours

Unit-III

Multi-Layer perceptron: Linear Inseparability: Hard problems, MLFFNN: Back propagation learning, Draw backs of back propagation algorithm, Heuristics to improve the performance of Back propagation learning discussion on error back propagation, Convolution neural network (CNN) 8 Hours

Unit-IV

Feedback Neural Networks: Analysis of pattern storage networks, The Hopfield Model, Energy analysis of Hopfield model, State transition diagram, Pattern storage: Hard problems, Stochastic Networks and simulated annealing.

Competitive learning network: Basic competitive learning, Analysis of pattern clustering Networks. Analysis of Feature Mapping Network 9 Hours

Unit-V

Architectures for complex pattern recognition tasks: Bidirectional associative memory, Architecture of Radial basis function (RBF) networks, Theorems for function approximation, RBF networks function approximation, Covers theorem on separability of patterns, The XOR problem, RBF Networks for pattern Classification, comparison of RBF with MLP networks. 8 Hours

Reference Books:

1 B. Yegnanarayana “ Artificial neural networks”, PHI, 2010.

2 Simon Haykin "Neural Networks for Pattern Recognition" Pearson Education Limited

3 Robert J. Schalkoff "Artificial Neural Networks”, Mcgraw-Hill Inc.

4 Jacek M. Zurada “Introduction to artificial neural systems”, Jaico publishing house, 2003.

5 Christopher M. Bishop “Neural networks for pattern recognition”, Oxford University Press (1995)