Abstract

This project aims to develop a robust system capable of predicting the prices of used cars based on various factors such as make, model, year, mileage, location, and condition. The rising demand for second-hand vehicles has led to the need for accurate pricing models, and this project utilizes machine learning techniques, particularly Artificial Neural Networks (ANNs), to address this challenge. The system is trained on a comprehensive dataset of used car listings, incorporating key features that impact car prices. Various machine learning algorithms, including linear regression, decision trees, and random forest, are tested and compared to assess their predictive accuracy. However, the core model relies on ANNs for its ability to capture complex, non-linear relationships in the data. By utilizing deep learning, the model can learn intricate patterns and make more accurate predictions, especially in cases where traditional models might struggle. The evaluation of the system is performed using standard regression metrics such as Mean Squared Error (MSE) and R-squared to ensure its reliability and performance. This predictive model not only provides a valuable tool for both buyers and sellers in the used car market but also demonstrates the potential of artificial intelligence in making data-driven decisions in the automotive industry.