Issue |
ITM Web Conf.
Volume 70, 2025
2024 2nd International Conference on Data Science, Advanced Algorithm and Intelligent Computing (DAI 2024)
|
|
---|---|---|
Article Number | 04032 | |
Number of page(s) | 8 | |
Section | AI and Advanced Applications | |
DOI | https://doi.org/10.1051/itmconf/20257004032 | |
Published online | 23 January 2025 |
Machine Learning Optimization and Challenges in Used Car Price Prediction
Santa Monica College, California, 90401, United States of America
Corresponding author: zheng_yufan01@student.smc.edu
With the rapid expansion of the second-hand vehicle market, correctly forecasting car prices is essential for both researchers and industry experts. The paper initially reviews existing machine learning models and their performance in predicting luxury car prices, emphasizing both their strengths and limitations. To begin with, models like XGBoost and Random Forest excel at processing large-scale data and identifying complex feature patterns, thanks to their ability to use an ensemble of decision trees to reduce bias and variance. However, these models struggle to accurately capture the unique characteristics of luxury vehicles, such as brand reputation, rarity, and personalized configurations. Because these complex factors cannot be easily represented by simple numerical features, the result is often suboptimal predictions for high-value vehicle prices. The paper found that feature engineering could enhance model performance by introducing more representative attributes specific to luxury vehicles, such as brand reputation, rarity, and customization options. Additionally, stratified modeling, which segments data based on price tiers, may provide more accurate predictions by targeting different price levels, especially in the high-value vehicle segment. Despite these theoretical benefits, the paper acknowledges that while these strategies were discussed, they were not empirically tested in detail. Consequently, their practical effectiveness still requires further investigation.
© The Authors, published by EDP Sciences, 2025
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.