Predict survival in the Titanic tragedy.
This is my current solution to the Titanic competition on Kaggle. I used Scikit-Learn to train, fine-tune and evaluate different classification models.
My current best model is a random forest classifier with these metrics:
- Training accuracy: 98.204%
- Test accuracy: 74.641%
- NumPy v0.13.2
- SciPy v1.2.1
- Pandas v0.24.1
- Joblib v0.13.2
- Kaggle 1.5.3
- Matplotlib v3.0.2
- Sklearn v0.20.3