Skip to content

54kennywang/ML_titanic

Repository files navigation

Kaggle Challenge: Titanic

This was my first ML project after taking the course Deep Learning Specialization on Coursera.

In the project, I mainly explored the area of different data processing and feature engineering techniques as in feature_engineering.py and Advanced_feature_engineering_refactored.py, including feature selection, missing data handling, categorical data to ordinal data conversion by mapping. Besides, I also studied the stacking technique as in stacking.py. stacking is defined with the specific example in stacking.py.

The highest accuracy rate was 0.80861, achieved using feature engineering with 5 layers of stacking. This ranks top 5% in the leadboard of the challenge.

Throughout this project, much of indepedent studies was done on the understading of different statistical concepts and various models, the summarys can be found here.

About

Kaggle Titanic challenge

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages