Predicting House Prices in King County Washington
Summary
The objective of this project was to build a model that can accurately predict house prices in King county Washington. In order to achieve this, Data Cleaning and feature engineering were performed to imporve the reliablilty of the results. This included Dimensionality reduction to reduce overfitting and Multicollinearity, standardization of the features and log transformations to vairbales that were rightly skewed. Once the final data table was created, the most curcial step of building the predictive model began. Several different algorithms were selected however Gradient Booting Regression proved to be the most accruate in terms of R Sqaured without overfitting the Data. A Grid Search with Cross Validation was used to find the optimal parameters for Gradient Boosting. Furthermore some notable obsverations were concluded such as the closer(in KM) the house is from downtown Seattle(Captial of Washington) the more expensive it is. The dataset used was taken from kaggle.