This is a small scale project which predicts the resale value of a car based upon several features like original price, kms driven, seller type, transition type, years used etc.
The sklearn.ensemble.RandomForestRegresser is used to train the model on a dataset from kaggle.
For the website, nodejs and express frameworks are used.
python3 (or python for Windows - Read Rectifying Errors)
node and npm
sklearn, pandas and numpy
Step 1: Download all the contents on this repo into your system
Step 2: Open your terminal and change its location to project's directory (cd path/Car-Resale-Value-Prediction/)
Step 3: Run "python3 price_prediction.py" (python price_prediction.py for Windows)
Step 4: Change your directory to Website/ (cd Website/)
Step 5: Run "npm start"
Step 6: Go to your browser and open localhost:3000
Step 7: Fill in the fields required and predict the value for your hypothetical car
This application was build on a Linux machine.
If you are running it on a windows machine, you will have to change python3 to python in app.js:20.
Delete model.sav in the main folder and execute the python script price_prediction.py
This will refresh the model that has been trained for predicting values (might take about 10-30 seconds)