<![CDATA[Sales is the most important component in and industrial company. This is because, income comes into the company if sales is running, therefore it is important to do an analysis of the products sold so that the company can prepare earlier before the demand for these products come, in order to produce better income. This study uses ARIMA, SVR, FFT and Prophet to forecast and with MAPE and RMSPE as the measure level of accuracy, also to see if there is any seasonality in the product that is being analyzed, seasonal_decompose is used. The results of the analysis show that ARIMA and Prophet are the best forecasting methods, this is because both methods have the lowest MAPE and RMSPE value. After being analyzed using seasonal_decompose, it was found that all of the products studied have a pattern that repeats itself at a certain time every quarter. For more further analysis, it was done by head-to-head comparison, where 20 product samples of each category was used. By this analysis it was clear that products in Category 1 are better to use ARIMA and products in Category 2 are better to use Prophet.]]>
Copyrights © 2023
