Solution for Kaggle Titanic Problem
https://www.kaggle.com/competitions/titanic
This file is for the Kaggle Machine learning problem predicting which passengers will survive the sinking fo the titanic. The code provides a solution to the problem that is 77% accurate to predicting which passengers will survive and which will perish. The best scores from an AI model are 80%. To improve this model, future work can change hyperparemeters or manipulate the data, especially regarding passengers with children and spouses, or you can decide that this model is good enough and work on other challenges.
This is a jupyter notebook and will work in jupyter or terminal, but it is really meant to be run in the Kaggle notebook. The notes in the code will provide instructions on how to submit to Kaggle for scoring but will only work if you are running the notebook in the Kaggle titianic competition linked above.
Of note: there are a number of 100% accurate predictions on the Kaggle leaderboard. These are achieved by manually determining which pasengers survive and perish from historical records and then hard coding their names into the solution. Approximately 80% is the best score achieved from machine learning techniques as of January 2023.