Ashleshk / Data-Science-Zillow-Real-Estate-Price-Prediction

Predict the price of unit area for houses given their features

Zillow Real Estate Price Prediction

Project Objectives

Every individual need to deal with the real estate or housing market at certain point of time in life. Having a good overview on the market will help in buying or selling the house in the market at right price.

In this project, Predict the price of unit area for houses given their features.

Data Preprocessing

• Drop Outliers - there were outliers in house price of unit area, longitude, distance to the nearest MRT station.
• Check Correlation
  • The number of convenience stores is moderately correlated to the price of unit area, while the distance to the nearest MRT station negatively correlated.
• Preprocessing Data
  • Convert transaction date to day, month and year columns
  • Scaling the data

Modelling

• Polynomial Regression
• XGBRegressor
• Ridge Regression
• Lasso Regression

Conclusion

1. R2 Score of the models

2. RMSE of the models

We can see that the XGBRegression Model performs the best

• R2 Score: 0.780957
• RMSE: 6.09142