Total views : 223

Analysis of Multiple Hidden Layer vs. Accuracy in Performance using Back Propagation Neural Network


  • CSE Department, Suresh Gyan Vihar University, Jagatpura, Jaipur, Rajasthan, India
  • SECE, Shri Mata Vaishno Devi University, Katra, Jammu - 182320, India


In this paper it has been analyzed that the proportionality among hidden layers neurons plays very important role in determining the accuracy of the target output. Better proportionality of neurons with appropriate number of hidden layers result in higher accuracy. In general, in any neural network at most two hidden layers are enough to train the network. But in some cases where accuracy is chief criteria then hidden layer plays vital role. This problem gets crucial when someone have to deal with multiscript numeral recognition where many numerals have similar shape but different values. For example ‘0’ shape in Arabic resembles numeric “Five” whereas in Hindi, English and many other scripting languages it resembles numeric “Zero”. This crucial problem usually arises when a person writes any number/pin code in two or more scripting language like Hindi and Farsi. These types of problems are taken into consideration and accuracy is considered chief criteria.


Multiple Hidden Layers, Multiscript Pin Code Recognition, Neural Network.

Full Text:

 |  (PDF views: 221)


  • Yang Z. Establishing structure for Artificial Neural Networks based on fractal. Journal of Theoretical and Applied Information Technology. 2013 Mar; 49(1).
  • Hegadi RS, Kamble PM. Recognition of handwritten Marathi numerals using multilayer feed forward Neural Network. 2014 World Congress on Computing and Communication Technologies; 2014.
  • Panchal G, Ganatra A, Kosta YP, Panchal D. Behavior analysis of multilayer perceptrons with multiple hidden neurons and hidden neurons. International Journal of Computer Theory and Engineering. 2011 Apr; 3(2). ISSN 1793-8201.
  • Kazuhiro S. A two phase method for determining the number of neurons in the hidden layer of a 3 layer Neural Network. 2010 SICE Annual Conference; The Grand Hotel, Taipei, Taiwan. 2010 Aug 18-21.
  • Lovassy R, KoczyLT, Gal L, Rudas IJ. Fuzzy Neural Networks stability in terms of the number of hidden layers. 12th IEEE International Symposium on Computational Intelligence and Informatics, CINTI 2011: Budapest, Hungary. 2011 Nov 21-22.
  • Rajesh Prasad J, Kulkarni UV. Gujrati character recognition using adaptive neuro fuzzy classifier. 2014 International Conference on Electronic Systems, Signal Processing and Computing Technologies; 2014.
  • Bhujade RK, Pandit A, Shiwani S. Improving accuracy in back propagation Neural Network by pruning proportionality in multiple hidden layers. Cienciae Tecnic aVitivinicola Journal. Portugal. 2016 Feb; 31(2). ISSN: 0254-0223.
  • Bhujade RK, Pandit A, Hemrajani N, Asthana S. A novel approach of five hidden layer proportionality in back propagation Neural Network. Fifth International Joint Conference on Advances in Engineering and Technology, AET; Kochi, India. 2014.
  • Bhujade RK, Pandit A. An analysis of effect of proportionality among hidden layers in standard back propagation Neural Network. International Journal of Computer Technology and Electronics Engineering. Florida. 2015 Aug; 5(4). ISSN 2249-6343.


  • There are currently no refbacks.

Creative Commons License
This work is licensed under a Creative Commons Attribution 3.0 License.