The algorithm id3 (Quinlan) uses the method top-down induction of decision trees. Given a set of classified examples a decision tree is induced, biased by the information gain measure, which heuristically leads to small trees. The examples are given in attribute-value representation. The set of possible classes is finite. Only tests, that split the set of instances of the underlying example languages depending on the value of a single attribute are supported.
標(biāo)簽：
algorithm
induction
decision
top-down
上傳時(shí)間：
2017-08-20
上傳用戶：hzy5825468
