Decision Tree Splitting Criterion

A Decision Tree Splitting Criterion is a metric function used by a decision tree algorithm to rank tree splitting options.

• Context:
  • It can range from being a Classification Tree Splitting Criterion, to being a Ranking Tree Splitting Criterion, to being a Regression Tree Splitting Criterion.
  • …
• Example(s):
  • a Classification Tree Splitting Criterion, such as: a Gini Index Impurity Function, Information Gain Impurity Function, or a Chi-Square Impurity Function.
  • a Ranking Tree Splitting Criterion.
  • a Regression Tree Splitting Criterion.
  • …
• Counter-Example(s):
  • a Decision Tree Pruning Algorithm.
• See: Splitting Criterion, Impurity Function.

---

---

References

2013

• http://en.wikipedia.org/wiki/Decision_tree_learning#Formulae
  • Algorithms for constructing decision trees usually work top-down, by choosing a variable at each step that best splits the set of items. Different algorithms use different metrics for measuring "best". These generally measure the homogeneity of the target variable within the subsets. Some examples are given below. These metrics are applied to each candidate subset, and the resulting values are combined (e.g., averaged) to provide a measure of the quality of the split.

1999

• (Shih, 1999) ⇒ Yu-Shan Shih. (1999). “Families of splitting criteria for classification trees].” In: Statistics and Computing, 9(4). doi:10.1023/A:1008920224518.

---