What is the learning of the tree decision -making?
Tree learning decision -making uses a predictive model with information branches similar to a tree to collect the assumptions and make a judgment on the value of the item. The system is used for machine learning, statistics and data mining. The decision -making trees are also known as regression or classification trees, depending on the purpose for which they are used. After reaching each element, whether through a computer or person, it must be determined whether it applies to the target item. Once each branch is explored, the answers can be used to determine the value.
In essence, the teaching of the tree is a process of answering questions. Each answer moves the process forward until there is enough information to make a decision. For example, a simple tree can begin by asking which of the two objects to buy. One question may ask if the object is useful, while the other may ask if one is better cost than the other. By asking all these questions it is usually possible to determine which actions Je Statistically advantageous. The answer to one question can lead to another. This could lead to some branches have many sub -branches, while others are less sophisticated because it is easy to answer the question. Performing the process in this way allows the user to develop a more detailed assessment of the item.
Another possible use of the teaching of the tree decision -making is categorization. Rather than having each question leads to a single decision, the set of information is divided into different areas based on the answer for each branch. Once all branches are categorized, the same process can also be started in each category.
Tree decision -making learning usually proceeds from the highest level down. He hates the flow of purchase. Once the question is fully answered, there is usually no need to refer to it until the results are assembled.
The results of the decision -making tree can be expressed in different ways.They may be the answer to the question yes or no or the number such as the price or time. The results can also detect the identity of an object and thus name the class in which it belongs.