Neural Network Models for Learning Patterns and Pattern Sequences

Sachin, H. B. (2017) Neural Network Models for Learning Patterns and Pattern Sequences. Masters thesis, Indian Institute of Science Education and Research Kolkata.

PDF (MS dissertation of Sachin H. B. (12MS048)) 12MS048.pdf - Submitted Version Restricted to Repository staff only Download (1MB)

Abstract

Neural networks are inspired from the biological neuronal networks which are modeled based on some properties of the neuron. These networks are capable of learning and having associative memory, which is of importance in pattern recognition, error correction, and time sequence retention to name a few. The associative memory property of one of the neural network, the hopfield network, has been studied and it is used to learn static patterns or memories. The problem of pattern sequence recognition has been studied using crosscorrelation matrix memory model. In both the cases the error correction capabilities and the memory of the network is studied through simulations.

Actions (login required)

View Item