Decision tree: concept, algorithm
Posted: Thu Feb 20, 2025 10:03 am
The idea of creating and further developing decision tree models appeared in the middle of the 20th century after the studies of probable human behavior by cyber systems. The works of K. Hoveland "Computer Modeling of Thinking" and E. Hunt "Experiments in Induction" played a leading role in the development of this direction.
Further increase in popularity of this method was provided by the works of John R. Quinlan, iceland mobile database who developed the ID3 algorithm and its improved modifications C4.5 and C5.0, as well as Leo Breiman, who proposed the CART algorithm and the random forest method.
In simple terms, a decision tree is a problem with several possible actions. The map shows the possible results of each step, as well as the reactions to them. This method is especially relevant in situations where you need to draw a conclusion about a series of sequential decisions leading to an optimal outcome.
Further increase in popularity of this method was provided by the works of John R. Quinlan, iceland mobile database who developed the ID3 algorithm and its improved modifications C4.5 and C5.0, as well as Leo Breiman, who proposed the CART algorithm and the random forest method.
In simple terms, a decision tree is a problem with several possible actions. The map shows the possible results of each step, as well as the reactions to them. This method is especially relevant in situations where you need to draw a conclusion about a series of sequential decisions leading to an optimal outcome.