ON HETEROGENOUS TRANSFER FUNCTIONS IN BAYESIAN NEURAL NETWORK

A Heterogeneous Bayesian Neural Network (HETBNN) model was derived with Gaussian error normal distribution, (). The prediction and model selection criterion is at different levels of hidden neurons and at different sample sizes. The hidden neurons include 2, 5, 10, while the sample sizes include 50, 100, 200, 500, 1000, and 5000. That is, taking each sample size, statistics were conducted at different levels of the choice of hidden neurons. The intention is to see the behavior of the network at different variables. Three primary transfer functions (HOTTFs), as well as two derived transfer functions (HETTFs) arising from the convolution of the HOTTFs, are namely; Symmetric Saturated Linear transfer function (SATLINS ), Hyperbolic Tangent transfer function (TANH), Hyperbolic Tangent sigmoid transfer function (TANSIG), Symmetric Saturated Linear and Hyperbolic Tangent transfer function (SATLINS-TANH) and Symmetric Saturated Linear and Hyperbolic Tangent Sigmoid transfer function (SATLINS-TANSIG). Simulation study was carried out in this work and Real life data are also used to establish the precision of the heterogeneous Bayesian models developed. The results showed that the HETFs performed better in terms of the forecast using Mean Square Error (MSE), Mean Absolute Error (MAE) and Test Error as the forecast prediction criteria.