a. Consider two possible splits of training records during the decision tree induction where Black and...

 

Transcribed Image Text:

a. Consider two possible splits of training records during the decision tree induction where Black and white colors represent class labels. i. A B Calculate the entropy of each split? Which split is more pure? ii. What is the information gain of each split? b. What strategies can help reduce overfitting in decision trees? a. Consider two possible splits of training records during the decision tree induction where Black and white colors represent class labels. i. A B Calculate the entropy of each split? Which split is more pure? ii. What is the information gain of each split? b. What strategies can help reduce overfitting in decision trees?