Approximating Number of Hidden layer neurons in Multiple Hidden Layer BPNN Architecture

Hidden layers plays a vital role in the performance of Back Propagation Neural Network especially in the case where problems related to arbitrary decision boundary to arbitrary accuracy with rational activation functions are encountered. Also, multiple hidden layer can approximate any smooth mapping to any accuracy. The process of deciding the number of hidden layers and number of neurons in each hidden layer is still confusing. In this paper, an survey is made in order to resolved the problem of number of neurons in each hidden layer and the number of hidden layers required