A General Framework for Learning from Weak Supervision

TL;DR

This paper introduces a versatile and scalable framework for weakly supervised learning that employs an EM formulation and automaton-based algorithms to handle diverse weak supervision sources efficiently.

Contribution

The paper presents a novel general framework (GLWS) for weak supervision that simplifies computation and broadens applicability across various weak supervision types.

Findings

Reduces computational complexity from quadratic or factorial to linear.

Demonstrates superior performance across 11 weak supervision scenarios.

Enhances scalability and versatility of weakly supervised learning.

Abstract

Weakly supervised learning generally faces challenges in applicability to various scenarios with diverse weak supervision and in scalability due to the complexity of existing algorithms, thereby hindering the practical deployment. This paper introduces a general framework for learning from weak supervision (GLWS) with a novel algorithm. Central to GLWS is an Expectation-Maximization (EM) formulation, adeptly accommodating various weak supervision sources, including instance partial labels, aggregate statistics, pairwise observations, and unlabeled data. We further present an advanced algorithm that significantly simplifies the EM computational demands using a Non-deterministic Finite Automaton (NFA) along with a forward-backward algorithm, which effectively reduces time complexity from quadratic or factorial often required in existing solutions to linear scale. The problem of learning from arbitrary weak supervision is therefore converted to the NFA modeling of them. GLWS not only enhances the scalability of machine learning models but also demonstrates superior performance and versatility across 11 weak supervision scenarios. We hope our work paves the way for further advancements and practical deployment in this field.