An analysis of the Kaggle dataset "Titanic: Machine Learning from Disaster" which can be found here: https://www.kaggle.com/c/titanic
This repo uses three different Machine Learning models to analyze the Titanic dataset. Each model uses the same preprocessor to shape the data prior to analysis, and uses the same training and testing dataset csvs.
This was a group project where a pair of programmers each created one model to compare results against each other using the same preprocessed data. The Kernel-SVM was the winner, by less than 1%. All models successfully classified ~90% of all entries.
- Neural Networks (1, 2, and 3) layers deep
- Kernel-SVM
- Random Forest using Ensemble learning