Optimization backpropagation algorithm based on Nguyen-Widrom adaptive weight and adaptive learning rate
Ulfi Andayani, Erna Budhiarti Nababan, Baihaqi Siregar, Muhammad Anggia Muchtar, Tigor Hamonangan Nasution, Ikhsan Siregar · 2017
The purpose of this research was to optimize the backpropagation algorithm process by adding the Nguyen-Widrow method in input layer of feed-forward process and adapting the learning rate parameter in backward process in the backpropagation. In the preprocessing usually the data have not been normalized so the significant to the target output need to be reduce in the input layer process [1]. By embedded Nguyen-Widrow method to adaptive the weight from the generate data that will be adapted by its weight. The changed value will influence the process in the feed-forward step [2]. If the value higher than the target output, the process will continue to adaptive the learning rate in the backward step. The value of the learning rate should be saliently large to allow a fast learning process but small enough to guarantee its effectiveness [3]. In the case of identification the type of a file based on the data input a sequence of training examples, the result of data testing obtained 92% accuracy.