My tiny repo for the House Prices competition on Kaggle. As always, I prefer to build on what others have done when it comes to ideas for data cleaning and preparation. Then I look into creating a couple of different models to train and test on the data.
Models that I used:
- Simple Linear Regression
- Ridge Regression
- ElasticNet Regression
- XGBRegressor
- Deep, Dense Neural Networks.
Based on the results, I suppose somewhat unsurprisingly, the XGB regressor performed the best. My model placed 548 out of 4585 submissions on Kaggle, based on the RMSE of the log of prediction vs. log of true value.