Predicting Automobile Prices Using Regression Methods and Particle Swarm Optimization

Abstract

This study examines automobile price prediction using three distinct regression methods: Ordinary Least Squares (OLS), Ridge Regression, and Particle Swarm Optimization (PSO). Utilizing data from the UCI Machine Learning Repository, we analyze the performance of these methods in modeling car prices based on engine size and horsepower. We compare OLS, a conventional linear regression technique, with Ridge Regression, which addresses multicollinearity through regularization, and PSO, a metaheuristic optimization method that efficiently handles non-linear relationships and complex data patterns. Performance is assessed using the Mean Squared Error (MSE), providing insights into the effectiveness of each approach. The findings illuminate the trade-offs between simplicity, regularization, and computational efficiency in regression modeling for car pricing, offering valuable insights for both industry practitioners and researchers.