Electronic Theses and Dissertation

Peramalan indeks harga perdagangan besar (ihpb) solar industri memerlukan metode yang mampu merepresentasikan ketidakpastian pada data runtun waktu. penelitian ini menerapkan metode fuzzy time series markov chain dengan penentuan interval menggunakan metode average based, sekaligus membandingkannya dengan penelitian nurhidayat (2024) yang menggunakan interval sturges. data yang digunakan merupakan data ihpb solar industri periode januari 2018 hingga desember 2024 yang bersumber dari situs resmi badan pusat statistik. untuk memperoleh perbandingan yang sebanding dengan penelitian sebelumnya, digunakan dataset 1 (periode januari 2018 hingga september 2023) yang terdiri atas 69 data, sedangkan dataset 2 (periode januari 2018 hingga desember 2024) terdiri atas 84 data. keakuratan hasil peramalan dievaluasi menggunakan mean absolute percentage error (mape). hasil penelitian menunjukkan bahwa metode fuzzy time series markov chain dengan interval average based menghasilkan nilai mape yang lebih kecil dibandingkan metode interval sturges. pada dataset 1, metode average based menghasilkan nilai mape sebesar 3,59%, sedangkan metode sturges menghasilkan nilai mape sebesar 5,44%. selanjutnya, pada dataset 2, metode average based menghasilkan nilai mape sebesar 3,21%, sedangkan metode sturges menghasilkan nilai mape sebesar 5,35%. hasil ini menunjukkan bahwa metode fuzzy time series markov chain dengan interval average based lebih akurat dibandingkan metode sturges. kata kunci: fuzzy time series markov chain, average based, dataset, peramalan, dan ihpb solar industri

Forecasting the Industrial Diesel Wholesale Price Index (WPI) requires a method capable of representing uncertainty in time series data. This study applies the Fuzzy Time Series Markov Chain method with interval determination using the Average Based method, while simultaneously comparing it with the research of Nurhidayat 2024 which used Sturges intervals. The data used is the Industrial Diesel WPI data for the period of January 2018 to December 2024, sourced from the official website of BPS-Statistics Indonesia. To obtain a comparable evaluation with previous research, Dataset 1 January 2018 to September 2023 consisting of 69 data points was used, while Dataset 2 January 2018 to December 2024 consists of 84 data points. The accuracy of the forecasting results was evaluated using Mean Absolute Percentage Error (MAPE). The results show that the Fuzzy Time Series Markov Chain method with Average Based intervals produces a smaller MAPE value compared to the Sturges interval method. In Dataset 1, the Average Based method produced a MAPE value of 3,59 percent, while the Sturges method produced a MAPE value of 5,44 percent. Furthermore, in Dataset 2, the Average Based method produced a MAPE value of 3,21 percent, while the Sturges method produced a MAPE value of 5,35 percent. These results indicate that the Fuzzy Time Series Markov Chain method with Average Based intervals is more accurate than the Sturges method. Keywords: Fuzzy Time Series Markov Chain, Average Based, Dataset, Forecasting, and WPI of Industrial Diesel