ACADSTAFF UGM

CREATION
Title : The effect of amplitude modification in S-shaped activation function on neural network regression
Author :

Faizal Makhrus, S.Kom., M.Sc., Ph.D. (1)

Date : 23 2023
Keyword : Neural network,activation functions,amplitude Neural network,activation functions,amplitude
Abstract : Time series forecasting using multilayer Feed-forward Neural Network (FNN) is potential to give high accuracy. Several factors influence the accuracy. One of them is the choice of activation functions (AFs). There are several AFs commonly used in FNN with their specific characteristics, such as bounded type AFs, namely sigmoid, softsign, arctan, and tanh. This paper investigates the effect of the amplitude in the bounded AFs in the FNN prediction accuracy. The investigation is in both theoretical and experimental. The theoretical analyses show that the high amplitudes give high accuracy when using softsign, arctan, and tanh AFs. Some conducted experiments support these theoretical analyses using foreign exchange datasets of 10 countries from different continents compared with US Dollar (USD). Based on the results, the optimum amplitude of the AFs should be high, which is greater or equal to 100 times the maximum input values of the FNN. https://doi.org/10.14311/NNW.2023.33.015
Group of Knowledge : Ilmu Komputer
Original Language : English
Level : Internasional
Status :
Published
Document
No Title Document Type Action
1 paper_cover_board.pdf
Document Type : [PAK] Full Dokumen
[PAK] Full Dokumen View
2 THE EFFECT OF AMPLITUDE MODIFICATION IN S-SHAPED ACTIVATION FUNCTION ON NEURAL NETWORK REGRESSION(1).pdf
Document Type : [PAK] Cek Similarity
[PAK] Cek Similarity View
3 Bukti_Scopus.pdf
Document Type : Dokumen Pendukung Karya Ilmiah (Hibah, Publikasi, Penelitian, Pengabdian)
Dokumen Pendukung Karya Ilmiah (Hibah, Publikasi, Penelitian, Pengabdian) View
4 Bukti_korespondensi_terurut_final.pdf
Document Type : [PAK] Bukti Korespondensi Penulis
[PAK] Bukti Korespondensi Penulis View