Structure Learning for Relational Logistic Regression: An Ensemble Approach

TL;DR

This paper introduces an ensemble learning algorithm for Relational Logistic Regression that uses functional-gradient boosting to efficiently learn vector-weighted first-order formulae, outperforming existing methods.

Contribution

It develops a novel boosting-based learning algorithm for RLR that simultaneously learns weights and formulas, improving efficiency and accuracy.

Findings

Outperforms existing RLR learning methods on standard datasets

Efficiently learns vector weights and formulas simultaneously

Demonstrates superior predictive performance

Abstract

We consider the problem of learning Relational Logistic Regression (RLR). Unlike standard logistic regression, the features of RLRs are first-order formulae with associated weight vectors instead of scalar weights. We turn the problem of learning RLR to learning these vector-weighted formulae and develop a learning algorithm based on the recently successful functional-gradient boosting methods for probabilistic logic models. We derive the functional gradients and show how weights can be learned simultaneously in an efficient manner. Our empirical evaluation on standard and novel data sets demonstrates the superiority of our approach over other methods for learning RLR.