WHAT-Datathon

WHAT datathon 22-23 April, Amsterdam! Enrique, Daniel and Aeron team

Pipeline

EDA

Data preprocessing

Cleaning, null removing, joining, concats, external info, prepare train, test and validation

Feature engineering

Main models

gSearch.ipynb -- can be used to find optimal parameter values for XGBoost models.

Model ensembling

stackEnsembler.ipynb -- can be used to stack ensemble multiple 1st level models from sklearn.

Interesting links

*Previous competitions, papers, forum entries, blog posts, ..