The purpose of this project is to try to implement a machine learning algrothim (Logistic Regression) from scratch and see how accuracte I can get without the benefit of any ML packages. I will be applying this algorithm on the infamous Titanic dataset in order to predict with some amount of accuracy whether individual passengers lived or died based on various features of the passenger. For this project, I will only be using the following libraries: Pandas, Numpy, MatPlotLib.
Tentative Game plan- will update later
- Read in the data
- Define functions necessary for logistic regression
- Use these functions on some very basic numeric columns
- Use learning curves to make sure everything is working
- Do some feature engineering
- Tune the final model
- See what the accuracy is
The data for this project was retreived from Kaggle.com/Titanic