Random Forest Classifier (From Scratch)

• Only using NumPy and SciPy, I built a Random Forest classifier using the Titanic dataset to predict passenger survival with 80% accuracy, based on features such as their position on the ship

Description

• This classifier also identified spam emails with 80% accuracy
• Utilized entropy as the purity function

One decision tree for the Titanic dataset

One decision tree for the Spam dataset

Validation accuracy vs. Maximum depth for the Spam dataset