Predictive learning via rule ensembles

TLDR

Regression and classification models are built as linear combinations of simple data‑derived rules. Each rule is a conjunction of a few simple statements about input variables, and the authors provide automated methods to identify interacting variables, quantify interaction strength, and visualize main and interaction effects. Rule ensembles achieve predictive accuracy comparable to state‑of‑the‑art methods while offering superior interpretability, as each rule is easily understood and its influence on predictions and variable relevance can be assessed globally, locally, or at individual points.

Abstract

General regression and classification models are constructed as linear combinations of simple rules derived from the data. Each rule consists of a conjunction of a small number of simple statements concerning the values of individual input variables. These rule ensembles are shown to produce predictive accuracy comparable to the best methods. However, their principal advantage lies in interpretation. Because of its simple form, each rule is easy to understand, as is its influence on individual predictions, selected subsets of predictions, or globally over the entire space of joint input variable values. Similarly, the degree of relevance of the respective input variables can be assessed globally, locally in different regions of the input space, or at individual prediction points. Techniques are presented for automatically identifying those variables that are involved in interactions with other variables, the strength and degree of those interactions, as well as the identities of the other variables with which they interact. Graphical representations are used to visualize both main and interaction effects.