Pengembangan Model Dynamic Pricing Untuk Penentuan Tarif Tiket Pada PT KAI Berbasis Peramalan Permintaan

Yunianto, Rahmanizar Maksum (2025) Pengembangan Model Dynamic Pricing Untuk Penentuan Tarif Tiket Pada PT KAI Berbasis Peramalan Permintaan. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6032231204-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (5MB) | Request a copy

Abstract

PT Kereta Api Indonesia (PT KAI) menghadapi tantangan dalam mengoptimalkan pendapatan di tengah keterbatasan kapasitas infrastruktur dan fluktuasi permintaan perjalanan penumpang. Penelitian ini bertujuan mengembangkan model dynamic pricing yang mampu menyesuaikan tarif tiket secara adaptif berdasarkan peramalan permintaan menggunakan pendekatan machine learning. Model peramalan dikembangkan dengan algoritma Long Short- Term Memory (LSTM) menggunakan data pemesanan tiket kereta Argo Bromo Anggrek dari H-45 hingga H-3 jam sebelum keberangkatan. Proses pengembangan mencakup tahap preprocessing, seasonal decomposition, normalisasi, serta tuning hyperparameter menggunakan Hyperband. Model terbaik menunjukkan performa peramalan yang layak dengan nilai MAPE sebesar 56,92% dan MAAPE sebesar 0,3183. Hasil peramalan digunakan dalam formulasi model dynamic pricing berbasis Nonlinear Integer Programming (NLIP), dengan mempertimbangkan kapasitas kereta, elastisitas harga, serta batasan tarif dan kenaikan harian. Simulasi dilakukan dengan pembulatan harga ke ribuan rupiah dan batas kenaikan Rp40.000 per hari. Hasil implementasi menunjukkan peningkatan rata-rata pendapatan harian sebesar 14% dibanding sistem tarif eksisting, dengan pertumbuhan tertinggi pada hari Kamis dan Sabtu. Penelitian ini membuktikan bahwa integrasi model LSTM dan NLIP dapat menjadi pendekatan strategis dalam sistem penetapan harga yang adaptif, efisien, dan berbasis data, yang berpotensi diterapkan lebih luas dalam sistem tarif dinamis PT KAI.

============================================================================================================================================

PT Kereta Api Indonesia (PT KAI) faces significant challenges in optimizing revenue amidst infrastructure limitations and fluctuating passenger demand. This study aims to develop a dynamic pricing model that adaptively adjusts ticket fares based on demand forecasting using a machine learning approach. The forecasting model was developed using the Long Short-Term Memory (LSTM) algorithm with historical booking data from H-45 to H-3 hours before train departure on the Argo Bromo Anggrek service. The development process included data preprocessing, seasonal decomposition, normalization, and hyperparameter tuning using the Hyperband method. The best-performing model demonstrated reliable forecasting capability with a MAPE of 56.92% and a MAAPE of 0.3183. Forecasting results were integrated into a dynamic pricing formulation using Nonlinear Integer Programming (NLIP), considering seat capacity, price elasticity, fare bounds, and a maximum daily price increment constraint. The simulation applied price rounding to the nearest thousand rupiah and a maximum daily increase of Rp40,000. Implementation results showed an average daily revenue increase of 14% compared to the existing fare system, with the highest gains observed on Thursdays and Saturdays. This research demonstrates that the integration of LSTM and NLIP can serve as a strategic, data-driven solution for adaptive and efficient fare setting, with potential application across PT KAI’s broader ticket pricing system.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Dynamic Pricing, Machine Learning, Long Short-Term Memory (LSTM), Nonlinear Integer Programming (NLIP), Peramalan. Dynamic Pricing, Machine Learning, Long Short-Term Memory (LSTM), Nonlinear Integer Programming (NLIP), Forecasting.
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. T Technology > T Technology (General) > T174 Technological forecasting T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T57.8 Nonlinear programming. Support vector machine. Wavelets. Hidden Markov models. T Technology > T Technology (General) > T57.83 Dynamic programming
Divisions:	Interdisciplinary School of Management and Technology (SIMT) > 61101-Master of Technology Management (MMT)
Depositing User:	Rahmanizar Maksum Yunianto
Date Deposited:	18 Jun 2025 06:02
Last Modified:	18 Jun 2025 06:02
URI:	http://repository.its.ac.id/id/eprint/119198

Actions (login required)

View Item