A New Technique to solve local Minima problem with large number of hidden nodes on Feed Forward Neural Network

  Kuldip Vora,  Shruti Yagnik

The Back-propagation (BP) algorithm is a well-known representative of all iterative gradient descent algorithms used for supervised learning in neural networks. It is its simplicity that attracts researchers and so that, many improvements and variations of the BP learning algorithm have been reported to beat its limitations such as slow convergence rate and convergence to the local minima. It is extensively used in many applications. In this paper we propose an algorithm that shows that by having structural changes we can still improve and can solve local minima problem with problems having large number of hidden nodes. Results have been shown on some classical problems such as parity problem and also on soil data classification problem

Keywords- Artificial neural network (ANN); Backpropagation algorithm(BPA); Hidden nodes; Target values; Local minima

Unique Identification Number - IJEDR1402111
Page Number(s) - 1978-1981
Pubished in - Volume 2 | Issue 2 | June 2014
DOI (Digital Object Identifier) -   
Publisher - IJEDR (ISSN - 2321-9939)

  Kuldip Vora,  Shruti Yagnik,   "A New Technique to solve local Minima problem with large number of hidden nodes on Feed Forward Neural Network", International Journal of Engineering Development and Research (IJEDR), ISSN:2321-9939, Volume.2, Issue 2, pp.1978-1981, June 2014, Available at :http://www.ijedr.org/papers/IJEDR1402111.pdf