Binary classification using Python on a publicly available dataset.
Budget: $20 – $60 CAD
Any dataset of your choice. Apply Random Forests and
two other classifiers of your choice. Train and test with a random split
(80% training & 20% testing). Calculate performance measures: sensitivity, specificity, and accuracy for
all three algorithms. Also, generate the ROC curve, and calculate the AUC score. For each algorithm, tune
the hyperparameters using grid search. Employ a feature selection method taught in the class on the training
dataset and re-run the same classification experiments. Output the performance measures, the ROC curve,
the AUC score before and after feature selection for all three algorithms.
two other classifiers of your choice. Train and test with a random split
(80% training & 20% testing). Calculate performance measures: sensitivity, specificity, and accuracy for
all three algorithms. Also, generate the ROC curve, and calculate the AUC score. For each algorithm, tune
the hyperparameters using grid search. Employ a feature selection method taught in the class on the training
dataset and re-run the same classification experiments. Output the performance measures, the ROC curve,
the AUC score before and after feature selection for all three algorithms.