Книга: Pruning (algorithm): Machine Learning, Decision Tree, Overfitting, Errors and Residuals in Statistics, Horizon Effect, Alpha-beta Pruning, Artificial Neural Network, Null-Move Heuristic

High Quality Content by WIKIPEDIA articles! Pruning is a technique in machine learning that reduces the size of decision trees by removing sections of the tree that provide little power to classify instances. The dual goal of pruning is reduced complexity of the final classifier as well as better predictive accuracy by the reduction of over fitting and removal of sections of a classifier that may be based on noisy or erroneous data.One of the questions that arises in a decision tree algorithm is the optimal size of the final tree. A tree that is too large risks overfitting the training data and poorly generalizing to new samples. A small tree might not capture important structural information about the sample space. In addition, it is impossible to tell if the addition of a single extra node will dramatically decrease error, a problem known as the horizon effect. A common strategy is to grow the tree until each node contains a small number of instances, perhaps two or five then use pruning to remove nodes that do not provide additional information.