Gå til indhold
Ordbog / Decision tree

Decision tree

Sidst opdateret 14. august 2023

What is a decision tree?

A decision tree is a popular algorithm that models decisions or actions as a tree-like flowchart structure. It is a predictive modeling technique used for both classification and regression tasks. Decision trees represent decisions and their possible consequences in a hierarchical structure, where each internal node represents a decision based on a specific feature or attribute, and each leaf node represents an outcome or prediction.

How are decision trees constructed?

Decision trees are constructed through a process called recursive partitioning, where the data is split based on specific criteria at each node to maximize information gain or minimize impurity. The criteria used for splitting can vary depending on the type of decision tree algorithms being used.

In a decision tree, the root node represents the initial decision or starting point. From the root node, branches extend to child nodes, which correspond to different possible outcomes or further decisions based on specific conditions. These conditions are determined by evaluating the values of different features or attributes. The tree branches continue to grow until reaching the leaf nodes, which provide the final predictions or outcomes.