PYTHON APPROACH ON FUZZY TIME SERIES ARIMA (1, 1, 1) MODEL TO ANALYSE ORIGINAL AND PREDICT RESULTS FOR ONLINE RETAIL OF FUEL BOOKING SERVICES

This paper contributes to modeling and forecasting gas booking demand in an online retail environment using time series techniques. Our work demonstrates how historical demand data can be utilized to estimate future demand and its impact on the supply chain. The historical demand data were used to create several autoregressive integrated moving average (ARIMA) models using the Box-Jenkins time series procedure. The best model was selected based on four performance criteria: statistical results, maximum likelihood, and standard error. The selected model, ARIMA (1, 1, 1), was validated using additional historical demand data under the same conditions. The results demonstrate that the model can effectively estimate and forecast future demand for gas booking in an online retail environment. These findings will provide trustworthy guidance to the company’s management in decision-making.