Formal Definition
Let denote a set of training examples, each of the form where is the value of the th attribute of example and is the corresponding class label. The information gain for an attribute is defined in terms of entropy as follows:
The information gain is equal to the total entropy for an attribute if for each of the attribute values a unique classification can be made for the result attribute. In this case the relative entropies subtracted from the total entropy are 0.
Read more about this topic: Information Gain In Decision Trees
Famous quotes containing the words formal and/or definition:
“It is in the nature of allegory, as opposed to symbolism, to beg the question of absolute reality. The allegorist avails himself of a formal correspondence between ideas and things, both of which he assumes as given; he need not inquire whether either sphere is real or whether, in the final analysis, reality consists in their interaction.”
—Charles, Jr. Feidelson, U.S. educator, critic. Symbolism and American Literature, ch. 1, University of Chicago Press (1953)
“Perhaps the best definition of progress would be the continuing efforts of men and women to narrow the gap between the convenience of the powers that be and the unwritten charter.”
—Nadine Gordimer (b. 1923)