Advanced Regression to Predict Housing Prices (Kaggle Competition)
Project Overview:
This project was a team effort along with Kweku Ulzen, Michael Chin, and Matthew Sun to enter into the Housing Prices in Ames, Iowa Kaggle Competition.
The goals included:
- Maximize the predictive power of our machine learning models to predict the sale prices of houses based on a given set of features.
- Code a feedback system that would allow the team to easily experiment with subtle changes to the vast variety of model tuning.
Approach:
The team set up a shared repository on Github, and set about initial dataset exploratory analysis independently, and sharing results frequently.
We came to a final model by following a cycle of data exploration, data engineering, feature selection, model selection and tuning, testing, and analyzing results.
Read our presentation here.
About Author
Dean Goldman
View all posts by Dean Goldman >Related Articles
Leave a Comment
No comments found.