Decision Trees Definition

Decision tree is a widely used data mining method. In decision theory, a decision tree is a graph of decisions and their possible consequences, represented in form of branches and nodes.

This data mining method is been used in various fields in business and science for many years and has given outstanding results.

Decision Tree Structure

Decision trees offer a symbolic decision-making model with high level of interpretability.

A decision tree is a special form of tree structure. The tree consists of nodes where a logical decision has to be made, and connecting branches that are chosen according to the result of this decision.

The nodes and branches that are followed constitute a sequential path through a decision tree that reaches a final decision in the end. (see our examples of decision trees for more information on their structure).

Decision Trees Creation

Decision trees are generated from the training data in a top-down direction. The root node of a decision tree is the trees initial state - the first decision node. Each node in a tree contains some data.

On a basis of an algorithm some calculations are completed, and the decision tree node is been split into two or more branches. In some cases, the node cannot be split, in this case it will be the final decision node.

The process is repeated until obtaining a completely discriminating tree. At this very point the decision tree might have nodes that are too specific to noise, that might be present in the training data. This is called over-fitting. To avoid over-fitting, a decision tree is generalized, by eliminating sub-trees.

Once a decision tree solution is generated from the learning data, it can be used for predictive analysis or estimating the best decision.

Application of a decision tree to a data example is a straightforward top-down decision-making process. The process can be controlled by starting it in the root node, taking the appropriate branch, and terminating when a leaf node is reached.

Decision trees can be created manually or with the help of data mining software.

Next >