The method of cut-off branches of the decision tree

When working with very large data sets to build a decision tree can be difficult. Even in cases where this is possible, it may not be the best way to use the data. As a workaround, use tree branches cut method decisions in the implementation phase. Pruning branches or replace some subtree branches held where this procedure does not lead to an increase in errors. The process takes place from the bottom up, thus it is rising. This procedure is more popular than the use of rules of a stop. Trees, received after the cut off some branches, called truncated. If a truncated tree still is not intuitive and hard to understand, using the extraction rules that combine to describe a set of classes. Each path from the root to its top or sheet gives one rule. Terms and Conditions are checks at internal nodes of the tree. The article describes the method of tree branches cut solutions to solve classification problems.