An Approach for Passengers Forecasting Using Fuzzy Time Series Undang Syaripudin, Wildan Budiawan Zulfikar, Wisnu Uriawan, Rifqi Ardian Nugraha
Department of ICT, Asia E-University, Malaysia
Department of Informatics, UIN Sunan Gunung Djati, Indonesia
LIRIS INSA de Lyon, France
Email: wildan.b[at]uinsgd.ac.id
Abstract
Nowadays, transportation is one of the main needs. A company engaged in public transportation can at least dispatch hundreds to thousands of passengers every day. At certain times the number of passengers is very difficult to predict. There was a time, when the number of passengers occurred a very significant surge. However, there are times when the number of passengers is drastically reduced. This is caused by many factors including time including the holiday season, holidays, and so on. As for other factors such as natural disasters, endemic diseases and others. The number of passengers must be proportional to the number of vehicles that have been prepared. The company must consider the condition of the vehicle and also the physical condition of the driver. The purpose of this study is to conduct passenger forecasting on the coming day using passenger data in the past using Fuzzy Time Series. This model has the accuracy or the difference between predictions with real data using MAD and MAPE that is equal to 34.6%.
Keywords: Fuzzy Time Series, Passenger, Forecasting
