This is a notebook that will take a full look at the Titanic Disaster data from the Kaggle competition linked here. I will follow my standard routine, that is:
- Defining the problem (Which Kaggle kindly did for us)
- Exploring the data
- Create initial model(s) and tuning the hyperparameters
- Evaluating the model(s)
- Experimenting untill satisfied
This notebook was created as sort of a test of the Google Colaboratory - most of the code was written in between patients at the dental facility where I am an intern. I managed to go through an entire workflow with data manipulation, model creation and then experimentation in the span of a few weeks (which is fast considering I had about 10-15 minute intervals for writing code).
For now I consider this notebook done. The predictions I made for the competition scored 76.555% which while not great is still satisfactory. I will probably return later to this repository to clean it up a bit, divide the notebook into separate files etc.
For now I'm out.
Ale-Cie, 21.03.2023