Analisis Peramalan Berat Angkutan Barang PT KAI Daerah Operasional 7 Madiun Menggunakan Variasi Kalender-ARIMA dan Long Short-Term Memory

Noliza, Anindya Putri (2026) Analisis Peramalan Berat Angkutan Barang PT KAI Daerah Operasional 7 Madiun Menggunakan Variasi Kalender-ARIMA dan Long Short-Term Memory. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2043221074-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Peningkatan dan fluktuasi volume angkutan barang kereta api pada periode tertentu memerlukan metode peramalan yang akurat untuk mendukung perencanaan operasional. Penelitian ini bertujuan membandingkan metode variasi kalender-ARIMA dan Long Short-Term Memory (LSTM) dalam meramalkan berat angkutan barang PT KAI Daerah Operasional 7 Madiun periode Januari 2022 hingga Agustus 2025. Penelitian ini bertujuan meramalkan berat angkutan barang PT KAI Daerah Operasional 7 Madiun menggunakan metode ARIMA dan Long Short-Term Memory (LSTM). Berdasarkan hasil penelitian dapat diketahui bahwa rata-rata berat angkutan sebesar 46.665 Kg, dengan nilai minimum 23.281 Kg dan maksimum 88.054 Kg. Hari raya Idul Fitri, Natal, dan Tahun Baru berpengaruh signifikan terhadap berat angkutan. Model ARIMA terbaik adalah (1,0,2)(0,1,1)7 dengan MAPE 21,90%, MAE 12.843,358, dan RMSE 14.700,907. Model LSTM terbaik menggunakan Batch Size 16, Hidden Layer 2, Neuron 100, Epoch 100, Learning Rate 0,01, dan time step 28, dengan MAPE 7,137%, MAE 3.351,967, dan RMSE 5.038,976. LSTM dinilai lebih baik untuk peramalan 181 periode, dengan hasil prediksi cenderung datar tanpa fluktuasi signifikan. Penelitian selanjutnya disarankan mengembangkan model dan variabel prediktor, sementara hasil penelitian dapat menjadi referensi serta bahan pertimbangan perencanaan operasional PT KAI Daop 7 Madiun.
===================================================================================================================================
The increasing and fluctuating volume of railway freight during certain periods requires accurate forecasting methods to support operational planning. This study aims to forecast the freight weight of PT Kereta Api Indonesia Operational Area 7 Madiun using Calendar Variation-ARIMA and Long Short-Term Memory (LSTM) methods based on data from January 2022 to August 2025. Based on the research findings, the average freight weight was 46,665 kg, with a minimum value of 23,281 kg and a maximum value of 88,054 kg. Eid al-Fitr, Christmas, and New Year significantly affected freight weight. The best ARIMA model was (1,0,2)(0,1,1)₇, with a MAPE of 21.90%, MAE of 12,843.358, and RMSE of 14,700.907. The best LSTM model used a batch size of 16, 2 hidden layers, 100 neurons, 100 epochs, a learning rate of 0.01, and a time step of 28, achieving a MAPE of 7.137%, MAE of 3,351.967, and RMSE of 5,038.976. LSTM performed better for forecasting 181 periods, with predictions showing a relatively flat pattern without significant fluctuations. Future research is recommended to develop more complex models and incorporate additional predictor variables, while the findings may serve as a reference and consideration for operational planning at PT KAI Operational Area 7 Madiun

Item Type:	Thesis (Other)
Uncontrolled Keywords:	ARIMA, Angkutan Barang, Long Short Term Memory, Peramalan, Variasi Kalender ARIMA, Calendar Variation, Forecasting, Freight Transportation, Long Short Term Memory
Subjects:	Q Science > QA Mathematics Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry)
Divisions:	Faculty of Vocational > 49501-Business Statistics
Depositing User:	Anindya Putri Noliza
Date Deposited:	05 Aug 2026 09:15
Last Modified:	05 Aug 2026 09:15
URI:	http://repository.its.ac.id/id/eprint/144090

Actions (login required)

View Item