Code for the Kaggle Titanic Project
1. Neural Network
2. XGBoost
3. Logistic Regression
My reflections on this project are that the cleaning matters as much as, or perhaps even more, than the training of the models. The pre-processing and cleaning of the data took the most amount of my time, and even then, i was not able to crack the top 50%.
This has made me realize the importance of cleaning data, even more so compared to my previous encounters with data science projects. Guess from here on out, i shall take the chance and opportunity to learn more about the different approaches and techniques there are to cleaning datasets.