Learning from interpretation transition using differentiable logic programming semantics

The combination of learning and reasoning is an essential and challenging topic in neuro-symbolic research. Differentiable inductive logic programming is a technique for learning a symbolic knowledge representation from either complete, mislabeled, or incomplete observed facts using neural networks....

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Machine learning Ročník 111; číslo 1; s. 123 - 145
Hlavní autoři: Gao, Kun, Wang, Hanpin, Cao, Yongzhi, Inoue, Katsumi
Médium: Journal Article
Jazyk:angličtina
Vydáno: New York Springer US 01.01.2022
Springer Nature B.V
Témata:
Algorithms
Artificial Intelligence
Computer Science
Control
Curricula
Datasets
Knowledge representation
Learning
Logic programming
Logic programs
Machine Learning
Mechatronics
Natural Language Processing (NLP)
Neural networks
Optimization
Robotics
Semantics
Simulation and Modeling
Special issue on Learning and Reasoning
Neuro-symbolic method
Differentiable inductive logic programming
Explainability
Learning from interpretation transition
Machine learning
ISSN:0885-6125, 1573-0565
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The combination of learning and reasoning is an essential and challenging topic in neuro-symbolic research. Differentiable inductive logic programming is a technique for learning a symbolic knowledge representation from either complete, mislabeled, or incomplete observed facts using neural networks. In this paper, we propose a novel differentiable inductive logic programming system called differentiable learning from interpretation transition (D-LFIT) for learning logic programs through the proposed embeddings of logic programs, neural networks, optimization algorithms, and an adapted algebraic method to compute the logic program semantics. The proposed model has several characteristics, including a small number of parameters, the ability to generate logic programs in a curriculum-learning setting, and linear time complexity for the extraction of trained neural networks. The well-known bottom clause positionalization algorithm is incorporated when the proposed system learns from relational datasets. We compare our model with NN-LFIT, which extracts propositional logic rules from retuned connected networks, the highly accurate rule learner RIPPER, the purely symbolic LFIT system LF1T, and CILP++, which integrates neural networks and the propositionalization method to handle first-order logic knowledge. From the experimental results, we conclude that D-LFIT yields comparable accuracy with respect to the baselines when given complete, incomplete, and mislabeled data. Our experimental results indicate that D-LFIT not only learns symbolic logic programs quickly and precisely but also performs robustly when processing mislabeled and incomplete datasets.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0885-6125
1573-0565
DOI:10.1007/s10994-021-06058-8