Motorcycles are everyone's choice as a means of transportation because they are affordable and can be used for a long time. The high level of motorcycle sales made CV Apollo Motor dealers experience difficulties in procuring motorcycle variants to be sold. The large number of motorcycle variants in one manufacturer makes sales different for each of these variants; there are variants with high and low sales. Therefore predictions about this matter are essential as information material for the company. Input data was obtained from CV Apollo Siantar from 2018 to 2022 as a sales prediction target consisting of 10 data based on Honda motorcycles. Each data has seven variables and one target. This data will later be transformed into data between 0 to 1 before training and testing are carried out using the Back-propagation algorithm artificial neural network. This study uses the back-propagation algorithm. Based on the analysis results, the best architectural model is 7-3-5-1 because it has the highest level of accuracy compared to other models, which is 100%. MSE Testing of 0.08501.
Copyrights © 2022