Suitable MLP network activation functions for breast cancer and thyroid disease detection

I. S. Isa, Z. Saad, S. Omar, M. K. Osman, K. A. Ahmad, H. A Mat Sakim

Research output: Chapter in Book/Report/Conference proceedingConference contribution

  • 19 Citations

Abstract

This paper presents a comparison study of various MLP activation functions for detection and classification problems. The most well-known (Artificial Neural Network) ANN architecture is the Multilayer Perceptron (MLP) network which is widely used for solving problems related to detection and data classifications. Activation function is one of the elements in MLP architecture. Selection of the activation functions in the MLP network plays an essential role on the network performance. A lot of studies have been conducted by reseachers to investigate special activation function to solve different kind of problems. Therefore, this paper intends to investigate the activation functions in MLP networks in terms of the accuracy performances. The activation functions under investigation are sigmoid, hyperbolic tangent, neuronal, logarithmic, sinusoidal and exponential. Medical diagnosis data from two case studies; thyroid disease and breast cancer, have been used to test the performance of the MLP network. The MLP networks are trained using Back Propagation learning algorithm. The performance of the MLP networks are calculated based on the percentage of correct classificition. The results show that the hyperbolic tangent function in MLP network had the capability to produce the highest accuracy for detecting and hence classifying breast cancer data. Meanwhile, for thyroid disease classification, neuronal function is the most suitable function that performed the highest accuracy in MLP network. © 2010 IEEE.

Original languageEnglish
Title of host publicationProceedings - 2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010
Pages39-44
Number of pages6
DOIs
StatePublished - 2010
Externally publishedYes
Event2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010 - Bali, Indonesia

Other

Other2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010
CountryIndonesia
CityBali
Period9/28/109/30/10

Fingerprint

Multilayer neural networks
Chemical activation
Hyperbolic functions
Network performance
Network architecture
Backpropagation
Learning algorithms
Neural networks

All Science Journal Classification (ASJC) codes

  • Computational Theory and Mathematics
  • Applied Mathematics
  • Modelling and Simulation

Cite this

Isa, I. S., Saad, Z., Omar, S., Osman, M. K., Ahmad, K. A., & Sakim, H. A. M. (2010). Suitable MLP network activation functions for breast cancer and thyroid disease detection. In Proceedings - 2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010 (pp. 39-44). [5701819] DOI: 10.1109/CIMSiM.2010.93

Suitable MLP network activation functions for breast cancer and thyroid disease detection. / Isa, I. S.; Saad, Z.; Omar, S.; Osman, M. K.; Ahmad, K. A.; Sakim, H. A Mat.

Proceedings - 2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010. 2010. p. 39-44 5701819.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Isa, IS, Saad, Z, Omar, S, Osman, MK, Ahmad, KA & Sakim, HAM 2010, Suitable MLP network activation functions for breast cancer and thyroid disease detection. in Proceedings - 2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010., 5701819, pp. 39-44, 2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010, Bali, Indonesia, 28-30 September. DOI: 10.1109/CIMSiM.2010.93
Isa IS, Saad Z, Omar S, Osman MK, Ahmad KA, Sakim HAM. Suitable MLP network activation functions for breast cancer and thyroid disease detection. In Proceedings - 2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010. 2010. p. 39-44. 5701819. Available from, DOI: 10.1109/CIMSiM.2010.93

Isa, I. S.; Saad, Z.; Omar, S.; Osman, M. K.; Ahmad, K. A.; Sakim, H. A Mat / Suitable MLP network activation functions for breast cancer and thyroid disease detection.

Proceedings - 2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010. 2010. p. 39-44 5701819.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

@inbook{959e7d82c9c64f7d9338d51fa7d79b3a,
title = "Suitable MLP network activation functions for breast cancer and thyroid disease detection",
abstract = "This paper presents a comparison study of various MLP activation functions for detection and classification problems. The most well-known (Artificial Neural Network) ANN architecture is the Multilayer Perceptron (MLP) network which is widely used for solving problems related to detection and data classifications. Activation function is one of the elements in MLP architecture. Selection of the activation functions in the MLP network plays an essential role on the network performance. A lot of studies have been conducted by reseachers to investigate special activation function to solve different kind of problems. Therefore, this paper intends to investigate the activation functions in MLP networks in terms of the accuracy performances. The activation functions under investigation are sigmoid, hyperbolic tangent, neuronal, logarithmic, sinusoidal and exponential. Medical diagnosis data from two case studies; thyroid disease and breast cancer, have been used to test the performance of the MLP network. The MLP networks are trained using Back Propagation learning algorithm. The performance of the MLP networks are calculated based on the percentage of correct classificition. The results show that the hyperbolic tangent function in MLP network had the capability to produce the highest accuracy for detecting and hence classifying breast cancer data. Meanwhile, for thyroid disease classification, neuronal function is the most suitable function that performed the highest accuracy in MLP network. © 2010 IEEE.",
author = "Isa, {I. S.} and Z. Saad and S. Omar and Osman, {M. K.} and Ahmad, {K. A.} and Sakim, {H. A Mat}",
year = "2010",
doi = "10.1109/CIMSiM.2010.93",
isbn = "9780769542621",
pages = "39--44",
booktitle = "Proceedings - 2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010",

}

TY - CHAP

T1 - Suitable MLP network activation functions for breast cancer and thyroid disease detection

AU - Isa,I. S.

AU - Saad,Z.

AU - Omar,S.

AU - Osman,M. K.

AU - Ahmad,K. A.

AU - Sakim,H. A Mat

PY - 2010

Y1 - 2010

N2 - This paper presents a comparison study of various MLP activation functions for detection and classification problems. The most well-known (Artificial Neural Network) ANN architecture is the Multilayer Perceptron (MLP) network which is widely used for solving problems related to detection and data classifications. Activation function is one of the elements in MLP architecture. Selection of the activation functions in the MLP network plays an essential role on the network performance. A lot of studies have been conducted by reseachers to investigate special activation function to solve different kind of problems. Therefore, this paper intends to investigate the activation functions in MLP networks in terms of the accuracy performances. The activation functions under investigation are sigmoid, hyperbolic tangent, neuronal, logarithmic, sinusoidal and exponential. Medical diagnosis data from two case studies; thyroid disease and breast cancer, have been used to test the performance of the MLP network. The MLP networks are trained using Back Propagation learning algorithm. The performance of the MLP networks are calculated based on the percentage of correct classificition. The results show that the hyperbolic tangent function in MLP network had the capability to produce the highest accuracy for detecting and hence classifying breast cancer data. Meanwhile, for thyroid disease classification, neuronal function is the most suitable function that performed the highest accuracy in MLP network. © 2010 IEEE.

AB - This paper presents a comparison study of various MLP activation functions for detection and classification problems. The most well-known (Artificial Neural Network) ANN architecture is the Multilayer Perceptron (MLP) network which is widely used for solving problems related to detection and data classifications. Activation function is one of the elements in MLP architecture. Selection of the activation functions in the MLP network plays an essential role on the network performance. A lot of studies have been conducted by reseachers to investigate special activation function to solve different kind of problems. Therefore, this paper intends to investigate the activation functions in MLP networks in terms of the accuracy performances. The activation functions under investigation are sigmoid, hyperbolic tangent, neuronal, logarithmic, sinusoidal and exponential. Medical diagnosis data from two case studies; thyroid disease and breast cancer, have been used to test the performance of the MLP network. The MLP networks are trained using Back Propagation learning algorithm. The performance of the MLP networks are calculated based on the percentage of correct classificition. The results show that the hyperbolic tangent function in MLP network had the capability to produce the highest accuracy for detecting and hence classifying breast cancer data. Meanwhile, for thyroid disease classification, neuronal function is the most suitable function that performed the highest accuracy in MLP network. © 2010 IEEE.

UR - http://www.scopus.com/inward/record.url?scp=79952087016&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79952087016&partnerID=8YFLogxK

U2 - 10.1109/CIMSiM.2010.93

DO - 10.1109/CIMSiM.2010.93

M3 - Conference contribution

SN - 9780769542621

SP - 39

EP - 44

BT - Proceedings - 2nd International Conference on Computational Intelligence, Modelling and Simulation, CIMSim 2010

ER -