My entries to Driven Data's tabular ML competition Richter's Predictor: Modeling Earthquake Damage.
As of 01/09/21 my best submission ranks 151 / 4593 - the top 3.5% of all submissions. This was achieved by bagging many LightGBM models together.
Note: 0/3 subs made refers to the number of submissions I am allowed to make each day.
I completed this project alongside a mentor to demonstrate my data science and machine learning skills. He paid me for it and left the following ⭐⭐⭐⭐⭐ review about working with me:
I paused work on this project in May 2021 and am in the process of tidying everything up so it can be presented to the world in a nice manner. You are one of the lucky souls who gets to see the repo in its raw form. But this means that not everything is as clean or orderly as it should be.
However, I hope it gives you an idea of how I approached this project and demonstrates my end-to-end data science skills using core libraries such as scikit-learn, pandas, numpy, and LightGBM on tabular data.
For this project, I used Python and the following libraries:
- Numpy
- Pandas
- Matplotlib
- Seaborn
- Scikit-learn
- LightGBM
- H2O.ai