Automated Machine Learning for Positive-Unlabelled Learning

TL;DR

This paper introduces two new Auto-ML systems, BO-Auto-PU and EBO-Auto-PU, designed for positive-unlabelled learning, and evaluates their performance against existing methods across diverse datasets.

Contribution

The paper presents two novel Auto-ML systems for PU learning based on Bayesian and evolutionary Bayesian optimisation, expanding the automation in PU classifier selection.

Findings

BO-Auto-PU and EBO-Auto-PU outperform existing methods on multiple datasets.

Extensive evaluation across 60 datasets demonstrates the effectiveness of the proposed Auto-ML systems.

EBO-Auto-PU shows superior performance in complex PU learning scenarios.

Abstract

Positive-Unlabelled (PU) learning is a growing field of machine learning that aims to learn classifiers from data consisting of labelled positive and unlabelled instances, which can be in reality positive or negative, but whose label is unknown. An extensive number of methods have been proposed to address PU learning over the last two decades, so many so that selecting an optimal method for a given PU learning task presents a challenge. Our previous work has addressed this by proposing GA-Auto-PU, the first Automated Machine Learning (Auto-ML) system for PU learning. In this work, we propose two new Auto-ML systems for PU learning: BO-Auto-PU, based on a Bayesian Optimisation approach, and EBO-Auto-PU, based on a novel evolutionary/Bayesian optimisation approach. We also present an extensive evaluation of the three Auto-ML systems, comparing them to each other and to well-established PU learning methods across 60 datasets (20 real-world datasets, each with 3 versions in terms of PU learning characteristics).