A Comparative Analysis of Machine Learning Methods for Estimating the Market Value of Used Cars
Теория и практика общественного развития · 27.05.2026 · (4) · 162-172
DOI: 10.24158/tipor.2026.4.20 ↗Kurz gefasst
Fachartikel in der Zeitschrift Теория и практика общественного развития, online erschienen am 27.05.2026. Verfasst von Alexander S. Chizhov und 2 weiteren Autor:innen.
Abstract (Originalsprache)
This article examines the problem of high-precision price forecasting in the Indian secondary market using ma-chine learning algorithms. The authors compare classical linear regression with ensemble methods: random forest and gradient boosting. The research methodology is based on a comprehensive machine learning ap-proach, including data cleaning and transformation, logarithmization of the target variable, forecasting based on step-by-step comparison of models by complexity, and business interpretation. Modern Python libraries such as Pandas, Numpy, Matplotlib, Seaborn, Scikit-learn, XGBoost, Joblib provide extensive capabilities for analyzing and forecasting large and heterogeneous data typical of the automotive market, as well as flexible customization and interactive visualization. The study argued for the need for a logarithmic transformation of the target variable to minimize data skewness. The results demonstrated the advantage of the gradient boost-ing model, yielding an R2 value of 0,94 and reducing the MAE absolute error to INR 1,387 lakh. A feature im-portance analysis was conducted, revealing the dominant role of engine power and year of manufacture in pricing.
Zitieren
APA
Chizhov, A. S., Mirolyubova, A. A., & Smirnova, O. P. (2026). A Comparative Analysis of Machine Learning Methods for Estimating the Market Value of Used Cars. Теория и практика общественного развития, 162-172. https://doi.org/10.24158/tipor.2026.4.20
Vancouver
Chizhov AS, Mirolyubova AA, Smirnova OP. A Comparative Analysis of Machine Learning Methods for Estimating the Market Value of Used Cars. Теория и практика общественного развития. 2026:162-172. doi:10.24158/tipor.2026.4.20
BibTeX
@article{Chizhov2026,
author = {Chizhov, Alexander S. and Mirolyubova, Anastasia A. and Smirnova, Olga P.},
title = {A Comparative Analysis of Machine Learning Methods for Estimating the Market Value of Used Cars},
journal = {Теория и практика общественного развития},
year = {2026},
number = {4},
pages = {162--172},
doi = {10.24158/tipor.2026.4.20}
}
EUVIMED