PREDICTION OF USED CAR PRICES USING ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING

Authors

  • DR.GOHIN Author
  • GUNNU DILEEP KUMAR Author

Keywords:

National Transport Authority, every six owns a car, price of second-hand cars using artificial  neural networks

Abstract

The number of cars on Mauritian roads has 
been rising consistently by 5% during the last 
decade. In 2014, 173 954 cars were registered 
at the National Transport Authority. Thus, 
one Mauritian in every six owns a car, most 
of which are second hand reconditioned cars 
and used cars. The aim of this study is to 
assess whether it is possible to predict the 
price of second-hand cars using artificial 
neural networks. Thus, data for 200 cars from 
different sources was gathered and fed to four 
different machine learning algorithms. We 
found that support vector machine regression 
produced slightly better results than using a 
neural network or linear regression. 
However, some of the predicted values are 
quite far away from the actual prices, 
especially for higher priced cars. Thus, more 
investigations with a larger data set are 
required and more experimentation with
required in order to obtain better predictions. 

Downloads

Published

16-04-2024

How to Cite

PREDICTION OF USED CAR PRICES USING ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING. (2024). International Journal of Engineering Research and Science & Technology, 20(2), 905-913. https://ijerst.org/index.php/ijerst/article/view/364