Neural Learning: Can we Make it a Little More Bio-Inspired!
Date of Submission
December 2018
Date of Award
Winter 12-12-2019
Institute Name (Publisher)
Indian Statistical Institute
Document Type
Master's Dissertation
Degree Name
Master of Technology
Subject Name
Computer Science
Department
Electronics and Communication Sciences Unit (ECSU-Kolkata)
Supervisor
Pal, Nikhil Ranjan (ECSU-Kolkata; ISI)
Abstract (Summary of the Work)
Though neural networks are inspired by human brains, recent studies have reported several differences between them in terms of their working principles. Specifically, some investigations on animal brains have shown that, in an animal brain, for different stimuli, different clusters neurons get activated. For example, when an animal visualizes different images, neurons from different parts of the brain gets activated. Being inspired from such an observation, here we propose a multilayered model of neural network that incorporates an idea of local activation of neurons for different group of objects (classes). In order to realize activation of distinct spatial clusters of neurons for different types of stimuli, the proposed model makes an interesting integration of a multi-layer perceptron and a self-organizing map. When compared to a conventional multilayer perceptron, the proposed model produces distinct locally activated regions for different classes and at the same time it learns to discriminate between classes.
Control Number
ISI-DISS-2018-395
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.
DOI
http://dspace.isical.ac.in:8080/jspui/handle/10263/6961
Recommended Citation
Bandyopadhyay, Ambar, "Neural Learning: Can we Make it a Little More Bio-Inspired!" (2019). Master’s Dissertations. 342.
https://digitalcommons.isical.ac.in/masters-dissertations/342
Comments
ProQuest Collection ID: http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqm&rft_dat=xri:pqdiss:28843401