/car-price-prediction

A car price prediction regression project to predict the price of cars. Includes exploratory data analysis, a linear model and random forest regressor.

Primary LanguageJupyter Notebook

Car Price Prediction

Objective:

To predict the price of the cars

Metric Used:

Root Mean Square Log Error(RMSLE)

(Ideally a value closer to 0.0 indicates a better performing model.)

Dataset:

  • The training dataset consisted of 19,236 rows and 18 columns. The test dataset consisted of 8,245 rows and 18 columns.
  • Exploratory Data Analysis - (Link to view code)

Method:

  1. Base-line model: Linear Regression Model (Code)
  • RMSLE on cross-validation dataset = 1.62
  • RMSLE on test dataset = 1.63
  1. Final Model: Random Forest Regressor (Code)
  • RMSLE on cross-validation dataset = 1.34
  • RMLSE on test dataset = 1.45

NOTE:

This project was as part of the MATHCO.THON hackathon hosted by MachineHack and organised by The Math Company. (View details)