This document discusses methods for predicting which passengers on the Titanic were most likely to survive using machine learning techniques. It analyzes data on passenger attributes like age, sex, class, and embarkation location. Random forests were selected for making predictions because they are flexible and easy to implement. Decision trees and entropy, which are components of random forests, are explained. The document tests different decision tree models and evaluates a random forest model from sklearn on the Titanic passenger data.