On PAC-Bayesian Bounds for Random Forests

TL;DR

This paper evaluates PAC-Bayesian bounds for random forests, showing that bounds based on Gibbs classifiers are often tighter than those considering ensemble correlations, with implications for understanding generalization performance.

Contribution

It compares different PAC-Bayesian approaches for bounding random forest errors, highlighting the advantages of Gibbs classifier bounds in correlated ensembles.

Findings

Gibbs classifier bounds are often tighter than C-bounds in correlated ensembles.

Correlation estimation can degrade the effectiveness of C-bounds.

Validation set bounds improve guarantees but reduce training data.

Abstract

Existing guarantees in terms of rigorous upper bounds on the generalization error for the original random forest algorithm, one of the most frequently used machine learning methods, are unsatisfying. We discuss and evaluate various PAC-Bayesian approaches to derive such bounds. The bounds do not require additional hold-out data, because the out-of-bag samples from the bagging in the training process can be exploited. A random forest predicts by taking a majority vote of an ensemble of decision trees. The first approach is to bound the error of the vote by twice the error of the corresponding Gibbs classifier (classifying with a single member of the ensemble selected at random). However, this approach does not take into account the effect of averaging out of errors of individual classifiers when taking the majority vote. This effect provides a significant boost in performance when the errors are independent or negatively correlated, but when the correlations are strong the advantage from taking the majority vote is small. The second approach based on PAC-Bayesian C-bounds takes dependencies between ensemble members into account, but it requires estimating correlations between the errors of the individual classifiers. When the correlations are high or the estimation is poor, the bounds degrade. In our experiments, we compute generalization bounds for random forests on various benchmark data sets. Because the individual decision trees already perform well, their predictions are highly correlated and the C-bounds do not lead to satisfactory results. For the same reason, the bounds based on the analysis of Gibbs classifiers are typically superior and often reasonably tight. Bounds based on a validation set coming at the cost of a smaller training set gave better performance guarantees, but worse performance in most experiments.