Suitable MLP Network Activation Functions for Breast Cancer and Thyroid Disease Detection

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