Questions

1. Apply the CSL technique to the SVM model from scikit-learn while utilizing the dataset that was used in this chapter. Use the class_weight and sample_weight parameters, similar to how we used them for other models in this chapter. Compare the performance of this model with the ones that we already encountered in this chapter.
2. LightGBM is another gradient-boosting framework similar to XGBoost. Apply the cost-sensitive learning technique to a LightGBM model while utilizing the dataset we used in this chapter. Use the class_weight and sample_weight parameters similar to how we used them for other models in this chapter. Compare the performance of this model with the ones that we already encountered in this chapter.
3. AdaCost [10] is a variant of AdaBoost that combines boosting with CSL. It updates the training distribution for successive boosting rounds by utilizing the misclassification cost. Extend AdaBoostClassifier from scikit-learn to implement the AdaCost algorithm...