Application of Backpropagation Neural Network in Predicting Mandatory Test Vehicle Parks

Abstract

Backpropagation neural networks can be used in almost every aspect of human life, including the prediction of mandatory test vehicle parks. The goal of this study was to use BPNN (Backpropagation Neural Network) modeling to anticipate mandatory test vehicle parks based on the past data from the Department of Transportation at Purwakarta Regency, and to predict the results using the best model. This study makes use of mandatory test vehicle parks data from 2014 through 2023, which necessitates monthly testing. The test results show an accuracy level of 90.134% utilizing alpha 0.9, iteration number (epoch) of 10000, and MSE value 0.0064. Based on the best BPNN model into Matlab applications, the mandatory test vehicle park will be predicted from June 2023 to May 2024. The estimated value of the mandatory test vehicle park in December 2023 will be used to determine the requirement for proof of passing the periodic test in 2023 with a score of 7587

Keywords

• backpropagation
• neural network
• mandatory test vehicle parks
• forecast
• mse