Decision Tree Concepts

Decision Tree Concepts

Over Fitting In a Decision Tree Context

Decision Trees are the most popular Data Mining tool. They are classifiers. They predict categorical outputs from categorical/real inputs. Over fitting in data is a significant issue in decision tree learning. There's been decades of research into refining it (e.g. dealing with continuous values and attributes with costs). When there are too many hypotheses that are consistent with the training set, we say that we are over?tting the training data. Over fitting denotes a certain correction of the model to a given data set (Mitchell,2004). In statistics, over fitting means ...