Predicting the survivability rate of breast cancer patients using Machine Learning

Courant Institute of Mathematical Sciences @ New York University

Big Data Science with Prof Anasse Bari

• In this project, we analyzed the Breast Cancer dataset and built machine learning models to predict the survival status of patients based on various features.
• We preprocessed the data by handling missing values, encoding categorical variables, identifying and removing outliers, and selecting relevant features.
• We implemented the KNN algorithm from scratch and evaluated the performance of the models using accuracy, precision, recall, and F1 score metrics.
• We compared the performance of different classifiers, including Linear Regression, Naive Bayes, Decision Tree, Random Forest, and Gradient Boosting, using hyperparameter tuning with GridSearchCV.
• Random Forest and Gradient Boosting classifiers achieved the highest recall, and F1 score on the test set, indicating that they are suitable for predicting the survival status of breast cancer patients.