Prediksi Harga Mobil Bekas Di Indonesia Dengan Pendekatan Algoritma Machine Learning Random Forest

Utomo, Rafiif Aliansyah Jaka (2024) Prediksi Harga Mobil Bekas Di Indonesia Dengan Pendekatan Algoritma Machine Learning Random Forest. Other thesis, Institute Teknologi Sepuluh Nopember.

Text 06311940000078-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (2MB) | Request a copy

Abstract

Revitalisasi bisnis mobil bekas melalui Revolusi Industri 4.0 memungkinkan transaksi lebih efisien, baik offline maupun online. Pertumbuhan pasar mobil bekas offline sejalan dengan mobilitas dan ekonomi, namun beberapa pelaku bisnis semakin menerapkan platform online. Platform seperti OLX, Carmudi, dan Mobil123 memberikan peluang bagi pembeli dan penjual untuk mencari, membandingkan dan mendapatkan informasi riwayat kendaraan. Tantangan tetap muncul dalam menentukan harga akurat karena variasi informasi di internet dan faktor-faktor harga yang kompleks. Pada penelitian ini dilakukan analisis prediksi harga jual dari mobil bekas yang ada di Indonesia dengan menggunakan algoritma random forest. Variabel yang digunakan adalah informasi mengenai mobil bekas yang didapatkan dari platform jual beli mobil bekas OLX, Carmudi, dan Mobil123 dengan periode data November 2018 sampai Desember 2022. Penelitian ini menghasilkan akurasi dari nilai error dan evaluasi lainnya yang dapat menjadi bahan pertimbangan untuk menggunakan model dan dapat memberikan wawasan mengenai faktor-faktor yang berpengaruh untuk menganalisis prediksi penghargaan mobil bekas berdasarkan variabel tertentu yang digunakan seperti variabel model mobil, variabel merek mobil, dan lain-lain. Hasil penelitian menunjukkan prediksi harga yang baik dengan nilai MAPE sekitar 4% untuk masing-masing merek dan keseluruhan merek. Berdasarkan hasil analisis menggunnakan metode mutual information juga didapatkan tiga variabel yang memiliki pengaruh paling signifikan terhadap prediksi harga, yaitu kapasitas mesin, tahun pembuatan mobil, dan model mobil.
=================================================================================================================================
The Fourth Industrial Revolution has transformed the paradigm of the used car business, enabling more efficient and transparent transactions, both conventionally and online. The growth of the offline used car market aligns with economic and mobility developments, yet dealers are increasingly adopting online platforms to reach customers from various regions. Additionally, online platforms like OLX, Carmudi, and Mobil123 offer opportunities for buyers and sellers to search, compare, and obtain comprehensive information about a vehicle's history. Determining the accurate price in used car transactions poses a challenge due to the diverse information available on the internet and unclear influencing factors. Therefore, having a reliable source of information to determine competitive prices is crucial in understanding the price range of cars in the market. In this study, an analysis of the predicted selling prices of used cars in Indonesia was conducted using the random forest machine learning algorithm. The variable used was the car information details obtained from the used car buying and selling platforms OLX, Carmudi, and Mobil123. This research aims to predict cases using machine learning algorithms and generate accuracy through error value assessments, providing insights into influential factors for analyzing predictions of used car prices based on specific variables. The research indicate accurate price predictions with an approximately 4% MAPE value for each brand and overall. The analysis also identifies three variables that most influence price predictions, namely engine capacity, year of manufacture, and car model, based on the calculation of the mutual information method.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Pasar Mobil Bekas, Prediksi, Machine Learning, Random Forest, Used Car Market
Subjects:	Q Science Q Science > QA Mathematics > QA9.58 Algorithms
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Actuaria > 94203-(S1) Undergraduate Thesis
Depositing User:	Rafiif Aliansyah Jaka Utomo
Date Deposited:	31 Jan 2024 02:04
Last Modified:	31 Jan 2024 02:04
URI:	http://repository.its.ac.id/id/eprint/105782

Actions (login required)

View Item