Open Access Open Access  Restricted Access Subscription or Fee Access

Significance of Activation Functions in Back Propagation Neural Network: A Study on Binary Categorical Data

S. Shanthi

Abstract


This work focuses on exploration of the performance measures of Back Propagation Neural Network (BPNN) learning algorithm using various activation functions and BPNN parameters viz. number of epochs, number of neurons and learning rate. The results show that the Tangent Sigmoid activation function with 20000 epochs, 15 neurons and 0.0001 learning rate has the capability to improve the accuracy of BPNN algorithm. The experiment is conducted on Diabetes dataset provided by the Department of Medicine, University of Virginia School of Medicine.

Keywords


Data Mining, Back propagation, Diabetes, Activation Function, Tangent Sigmoid, Epoch, Neurons, Learning Rate.

Full Text:

PDF

References


Aurangieb Khan and Kenneth Revett, ” Data Mining the PIMA Dataset Using Rough Set Theory with a Special Emphasis on Rule Reduction”, 2004, Multitopic Conference and Proceedings of INMIC, pp. 334 - 339

Bekir Karlik, and A Vehbi Olgac, ” Performance Analysis of Various Activation Functions in Generalized MLP Architectures of Neural Networks”,2010, International Journal of Artificial Intelligence and Expert Systems (IJAE), Vol. 1(4) ,pp. 111-122.

Coskun Özkan and Filiz Sunar Erbek,”The Comparison of Activation Functions for Multispectral Landsat TM Image Classification”, 2003, Photogrammetric Engineering & Remote Sensing, Vol. 69, No. 11, pp. 1225–1234.

Emad A. M. Andrews Shenouda,” A Quantitative Comparison of Different MLP Activation Functions in Classification”, 2006, Proceedings of the Third international conference on Advances in Neural Networks, Vol.1.

Farhad Soleimanian Gharehchopogh, Zeynab Abbasi Khalifelu,”Neural Network Application in Diagnosis of Patient: A Case Study”, 2011,Computer Networks and Information Technology (ICCNIT), pp. 245 – 249.

Hian Chye Koh and Gerald Tan, “Data Mining Applications in Healthcare”, 2003, Journal of Healthcare Information Management, Vol. 19, No. 2.

Hongjun Lu, Rudy Setiono, and Huan Liu,” Effective Data Mining Using Neural Networks”, 1996,IEEE Transactions on knowledge and Data mining , Vol. 8, Issue-6, pp- 957 – 961.

I.S.Isa, Z.Saad, S.Omar, M.K.Osman, K.A.Ahmad and H.A.Mat Sakim,” Suitable MLP Network Activation Functions for Breast Cancer and Thyroid Disease Detection”, 2010, Second International Conference on Computational Intelligence, Modeling and Simulation, pp- 39 - 44.

Jayalskshmi and Dr.A.Santhakumaran,” Impact of Preprocessing for Diagnosis of Diabetes Mellitus Using Artificial Neural Networks”, 2010, Second International Conference on Machine Learning and Computing, pp- 109 - 112.

Jiawei Han and Micheline Kamber,” Data Mining: Concepts and Techniques”.

Manaswini Pradhan and Dr. Ranjit Kumar Sahu,” Artificial Neural Network (ANN) Technology and Disease Surveillance: A Study on Diabetes”, 2011, International Journal of Research and Reviews in Computer Science (IJRRCS) Vol. 2, No. 2.

Mehdi Rezaeian Zadeh, Seifollah Amin, Davar Khalili, Vijay P. Singh, ” Daily Outflow Prediction by Multi Layer Perceptron with Logistic Sigmoid and Tangent Sigmoid Activation Functions”,2010, Water Resources Management, Vol.24, Number 11, pp. 2673-2688(16).

Pravin Chandra, Yogesh Singh,” An Activation function adapting training algorithm for sigmoidal feed forward networks”, 2004, Elsevier- Neuro computing, Vol 61, pp.429 – 437.

Rahul Kala, Anupam Shukla, Ritu Tiwari, ” Comparative Analysis of Intelligent Hybrid Systems for detection of PIMA Indian Diabetes”, 2009,World Congress on Nature & Biologically Inspired Compusting (NaBIC), pp- 947 - 952.

S.N Sivanandam, S.N.Deepa,” Principles of Soft Computing”, Second Edition, Wiley India Edition.

Wei Xu, Tingting Zheng, Ziang Li, ” A Neural Network Based Forecasting Method for the Unemployment Rate Prediction Using the Search Engine Query Data” , 2011,IEEE International Conference on e-Business Engineering, pp.9 - 15 .

William J.Palm,”Introduction to Matlab 7.4”, First Edition, Tata McGraw Hill.


Refbacks

  • There are currently no refbacks.