Towards Probabilistic Inductive Logic Programming with Neurosymbolic Inference and Relaxation

Many inductive logic programming (ILP) methods are incapable of learning programs from probabilistic background knowledge, for example, coming from sensory data or neural networks with probabilities. We propose Propper, which handles flawed and probabilistic background knowledge by extending ILP wit...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Theory and practice of logic programming Ročník 24; číslo 4; s. 628 - 643
Hlavní autori: HILLERSTRÖM, FIEKE, BURGHOUTS, GERTJAN
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Cambridge, UK Cambridge University Press 01.07.2024
Predmet:
Artificial intelligence
Background noise
Deep learning
Graph neural networks
Hypotheses
Hypothesis testing
Knowledge
Learning programs
Logic programming
Methods
Neural networks
Original Article
Probabilistic inference
Probability
Statistical analysis
Statistical models
neurosymbolic inference
probabilistic background knowledge
sensory data
relational patterns
inductive logic programming
ISSN:1471-0684, 1475-3081
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Many inductive logic programming (ILP) methods are incapable of learning programs from probabilistic background knowledge, for example, coming from sensory data or neural networks with probabilities. We propose Propper, which handles flawed and probabilistic background knowledge by extending ILP with a combination of neurosymbolic inference, a continuous criterion for hypothesis selection (binary cross-entropy) and a relaxation of the hypothesis constrainer (NoisyCombo). For relational patterns in noisy images, Propper can learn programs from as few as 8 examples. It outperforms binary ILP and statistical models such as a graph neural network.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1471-0684
1475-3081
DOI:10.1017/S1471068424000371