Project 3: Machine Learning Prediction of the Titanic Disaster
-
Created a tool that predicted the survival of passengers based on various features such as age, gender, and ticket class using machine learning models.
-
Cleaned and preprocessed data using methods such as, imputing missing values, encoding categorical features, and scaling numeric features.
-
Engineered features from the data to model using logistic regression, decision trees, random forests, and support vector machines (SVMs) algorithms.
-
Optimized the algorithms using GridSearchCV to reach best model.
-
Evaluated the models using metrics such as, accuracy, precision, recall, and F1-score. The random forest model performed the best with an accuracy of 82%.
-
The insights gained from the project can be useful in understanding the factors that contributed to survival on the Titanic and can inform future disaster preparedness efforts.
