Forecasting the Cost of Resale Motorcars

Forecasting the Cost of Resale Motorcars

Sanjay K S 1,Sindhu S L 2 1
Student,4th Semester MCA, Department of MCA, BIET,

Davanagere 2 Assistant Professor, Department of MCA, BIET, Davanagere

Abstract:

The increasing popularity of cars has led to a situation where many potential buyers are unable to purchase new vehicles due to factors such as high costs and limited availability. This has fueled the growth of the pre-owned car market worldwide. However, in India, this market is still developing and largely controlled by unorganized sectors, increasing the risk of fraud for buyers. Therefore, there is a need for a precise and unbiased model to estimate the price of pre-owned cars, benefiting both customers and sellers. This project focuses on developing a supervised learning-based Artificial Neural Network (ANN) model and a Random Forest machine learning model to predict pre-owned car prices using a given dataset. The aim is to create a working model with low error. The study examines various attributes to ensure reliable and accurate price prediction. The results demonstrate improvements over simpler linear models. An ANN model is built using the Keras Regressor algorithm, and other machine learning algorithms, including Random Forest, Lasso, Ridge, and Linear Regression, are also implemented. These algorithms are tested using the car dataset. The experimental results show that the Random Forest model achieves the lowest error, with a Mean Absolute Error (MAE) of 1.0970472 and an R-squared value of 0.772584. This work highlights the potential of using Random Forest for pre-owned car price prediction and suggests avenues for future research to further reduce fraud in this market, potentially achieving complete accuracy in the future..

Keywords: Pre-owned car price prediction, supervised learning, artificial neural network (ANN), Random Forest, Keras Regressor, machine learning, regression algorithms, fraud detection, India, automotive market, price estimation, model evaluation, mean absolute error (MAE), R-squared, dataset analysis.