Abstract:Decision trees are popular classification models, providing high accuracy and intuitive explanations. However, as the tree size grows the model interpretability deteriorates. Traditional tree-induction algorithms, such as C4.5 and CART, rely on impurity-reduction functions that promote the discriminative power of each split. Thus, although these traditional methods are accurate in practice, there has been no theoretical guarantee that they will produce small trees. In this paper, we justify the use of a general family of impurity functions, including the popular functions of entropy and Gini-index, in scenarios where small trees are desirable, by showing that a simple enhancement can equip them with complexity guarantees. We consider a general setting, where objects to be classified are drawn from an arbitrary probability distribution, classification can be binary or multi-class, and splitting tests are associated with non-uniform costs. As a measure of tree complexity, we adopt the expected cost to classify an object drawn from the input distribution, which, in the uniform-cost case, is the expected number of tests. We propose a tree-induction algorithm that gives a logarithmic approximation guarantee on the tree complexity. This approximation factor is tight up to a constant factor under mild assumptions. The algorithm recursively selects a test that maximizes a greedy criterion defined as a weighted sum of three components. The first two components encourage the selection of tests that improve the balance and the cost-efficiency of the tree, respectively, while the third impurity-reduction component encourages the selection of more discriminative tests. As shown in our empirical evaluation, compared to the original heuristics, the enhanced algorithms strike an excellent balance between predictive accuracy and tree complexity.

Unifying Attribute Splitting Criteria of Decision Trees by Tsallis Entropy.

Unifying Decision Trees Split Criteria Using Tsallis Entropy

Rough set based decision tree

Improving Decision Trees by Tsallis Entropy Information Metric Method.

A less-greedy two-term Tsallis Entropy Information Metric approach for decision tree classification

Dynamic Split-point Selection Method for Decision Tree Evolved by Gene Expression Programming

A New Approach of the Split Criteria of Decision Trees Based on the T-entropy

Statistical analysis of various splitting criteria for decision trees

Evidential Decision Tree Based on Belief Entropy

Heuristic Mode Research and Application of Decision Tree Algorithm

An improved decision tree algorithm based on boundary mixed attribute dependency

Harnessing the Power of Choices in Decision Tree Learning

Heuristic Mode Research Of Decision Tree Algorithm Based On Synthesizing Effect

The Power of Unbiased Recursive Partitioning: A Unifying View of CTree, MOB, and GUIDE

Building decision trees for the multi-class imbalance problem

Significance-Based Decision Tree for Interpretable Categorical Data Clustering

Correlation and Unintended Biases on Univariate and Multivariate Decision Trees

Analyzing CART

Learning accurate and interpretable decision trees

Regularized impurity reduction: Accurate decision trees with complexity guarantees

Permutation Decision Trees