Predicting the Number of Train Passengers in Java Island using SARIMA Model

Luluk Muthoharoh (Data Science Study Program, Science Department, Sumatra Institute of Technology)
Dimas Wahyu Saputro (Data Science Study Program, Science Department, Sumatra Institute of Technology)
Dhea Sukma Agustiana (Data Science Study Program, Science Department, Sumatra Institute of Technology)
Fadia Dilla Sabine (Data Science Study Program, Science Department, Sumatra Institute of Technology)
Lis Nuraini (Data Science Study Program, Science Department, Sumatra Institute of Technology)
Rekzi P. Manullang (Data Science Study Program, Science Department, Sumatra Institute of Technology)
Taj Shavira (Data Science Study Program, Science Department, Sumatra Institute of Technology)
Mika Alvionita (Data Science Study Program, Science Department, Sumatra Institute of Technology)

EKSAKTA: Journal of Sciences and Data Analysis
VOLUME 4, ISSUE 2, August 2023

The train at this time has become one of the most popular public transportation for medium and long-distance travel. The number of train passengers is difficult to predict during the holiday season. This study aimed to predict the number of train passengers using the Seasonal Autoregressive Integrated Moving Average (SARIMA) method. The stages used in this study include (1) dataset preparation, (2) preprocessing data, and (3) experimental testing and methods. The SARIMA model obtained is ARIMA(2,1,0)(0,1,2)[12] with an AIC value of 2379,265. A diagnostic model was carried out, and it was found that the model is quite good. So the SARIMA method used in predicting passengers is accurate.