PREDIKSI KEUNTUNGAN EKSPOR DENGAN METODE FUZZY TIME SERIES MODEL MARKOV CHAIN (STUDI KASUS : PROVINSI KEPULAUAN RIAU)

Abstract

Kepulauan Riau is one of the export-producing Provinces in Indonesia. Basically, export profits do not always increase as expected. At a certain time, export profits can decrease or even suffer losses. Based on the problems above, a system that can predict export profits is needed. Many methods can be used to make predictions. In this study, the author will predict the export profit in the Kepulauan Riau using the Fuzzy Time Series Model Markov Chain method. The data used is export profit data from January 2014 to April 2021 in the Kepulauan Riau Province. The predicted data will then be calculated for its accuracy using Mean Absolute Percentage Error (MAPE). Tests were carried out 3 times using positive numbers D1 and D2. The first test uses positive number input D1 = 40,000 and D2 = 900,000,000, the second test uses positive number input D1 = 600,000,000 and D2 = 100,000,000, and the third test uses positive number input D1 = 15,000,000 and D2 = 30.000. The prediction results obtained with the smallest error rate in the second test where MAPE is 7,8375 %