PENERAPAN METODE DESEASONALIZED PADA PERAMALAN BANYAK PENUMPANG KERETA API DI PULAU JAWA

Abstract

Forecasting is an estimates of some phenomenon that will be occurred
future time. Forecasting uses past data that will be analyzed scientifically.
Deseasonalized is a part of decomposition methods that be included to
time series methods. The base concept of this methods is clear all effects of
seasonal movement or variation, that possibly make us focus on long term
movement. The aim of this research is knowing the form of deseasonalized
models and forecasting result of how much the train passenger in Java
Island at year 2017 and 2018. This forecasting done by counting seasonal
index using ratio methods of moving average, and then determine the
equation of the trends. The forecasting calculated by multiplying Y
(estimated value) and seasonal index. From this research, we get the
deseasonalized model Y=40165,28+2589,163t. The forecasting result
of the number of train passenger in Java Island for first quarters at 2017
until forth quarters at 2018 in consecutive series are 90700, 97001,
101548, 103688, 100636, 107344, 112095, 114185.