Salvato in:
Dettagli Bibliografici
Autori principali: Wang, Kaifu, Tsamoura, Efthymia, Roth, Dan
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:https://arxiv.org/abs/2306.13796
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916321179992064
author Wang, Kaifu
Tsamoura, Efthymia
Roth, Dan
author_facet Wang, Kaifu
Tsamoura, Efthymia
Roth, Dan
contents We consider a weakly supervised learning scenario where the supervision signal is generated by a transition function $σ$ of labels associated with multiple input instances. We formulate this problem as \emph{multi-instance Partial Label Learning (multi-instance PLL)}, which is an extension to the standard PLL problem. Our problem is met in different fields, including latent structural learning and neuro-symbolic integration. Despite the existence of many learning techniques, limited theoretical analysis has been dedicated to this problem. In this paper, we provide the first theoretical study of multi-instance PLL with possibly an unknown transition $σ$. Our main contributions are as follows. Firstly, we propose a necessary and sufficient condition for the learnability of the problem. This condition non-trivially generalizes and relaxes the existing small ambiguity degree in the PLL literature, since we allow the transition to be deterministic. Secondly, we derive Rademacher-style error bounds based on a top-$k$ surrogate loss that is widely used in the neuro-symbolic literature. Furthermore, we conclude with empirical experiments for learning under unknown transitions. The empirical results align with our theoretical findings; however, they also expose the issue of scalability in the weak supervision literature.
format Preprint
id arxiv_https___arxiv_org_abs_2306_13796
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle On Learning Latent Models with Multi-Instance Weak Supervision
Wang, Kaifu
Tsamoura, Efthymia
Roth, Dan
Machine Learning
We consider a weakly supervised learning scenario where the supervision signal is generated by a transition function $σ$ of labels associated with multiple input instances. We formulate this problem as \emph{multi-instance Partial Label Learning (multi-instance PLL)}, which is an extension to the standard PLL problem. Our problem is met in different fields, including latent structural learning and neuro-symbolic integration. Despite the existence of many learning techniques, limited theoretical analysis has been dedicated to this problem. In this paper, we provide the first theoretical study of multi-instance PLL with possibly an unknown transition $σ$. Our main contributions are as follows. Firstly, we propose a necessary and sufficient condition for the learnability of the problem. This condition non-trivially generalizes and relaxes the existing small ambiguity degree in the PLL literature, since we allow the transition to be deterministic. Secondly, we derive Rademacher-style error bounds based on a top-$k$ surrogate loss that is widely used in the neuro-symbolic literature. Furthermore, we conclude with empirical experiments for learning under unknown transitions. The empirical results align with our theoretical findings; however, they also expose the issue of scalability in the weak supervision literature.
title On Learning Latent Models with Multi-Instance Weak Supervision
topic Machine Learning
url https://arxiv.org/abs/2306.13796