Logistic Regression to predict benign vs malignant tumors.
- Pandas
- Numpy
- Matplotlib
- Seaborn
- Scikit-Learn
The objective of the project is to explore the sampled breast cancer data from the Wisconsin Breast Cancer database and predict if a tumor is benign or malignant. The data mainly describes the properties of the cell and its class. The model used for the prediction is a logistic regression regularized by L1 to avoid over-fitting. The best features for the regression were selected by using a Recursive Feature Elimination model which given the logistic regression model assigns weights to the features (predictors) to select the features with higher importance.