which are correct

• A decision tree is learned by maximizing information gain.
• To predict the response for a given observation in a classification tree, we typically use the mean of the training observations in the region to which it belongs.
• Recursive binary splitting is a bottom-up greedy approach that can be used to find the best split at every step.
• A pruned decision tree might lead to lower variance and better interpretation, at the cost of a little bias.
• When generating a classification tree, a small value of the Gini index indicates a higher node purity.
• When the true decision boundary is linear, decision trees tend to perform better than linear regression.