Learning explanatory rules from noisy data
From MaRDI portal
Abstract: Artificial Neural Networks are powerful function approximators capable of modelling solutions to a wide variety of problems, both supervised and unsupervised. As their size and expressivity increases, so too does the variance of the model, yielding a nearly ubiquitous overfitting problem. Although mitigated by a variety of model regularisation methods, the common cure is to seek large amounts of training data---which is not necessarily easily obtained---that sufficiently approximates the data distribution of the domain we wish to test on. In contrast, logic programming methods such as Inductive Logic Programming offer an extremely data-efficient process by which models can be trained to reason on symbolic domains. However, these methods are unable to deal with the variety of domains neural networks can be applied to: they are not robust to noise in or mislabelling of inputs, and perhaps more importantly, cannot be applied to non-symbolic domains where the data is ambiguous, such as operating on raw pixels. In this paper, we propose a Differentiable Inductive Logic framework, which can not only solve tasks which traditional ILP systems are suited for, but shows a robustness to noise and error in the training data which ILP cannot cope with. Furthermore, as it is trained by backpropagation against a likelihood objective, it can be hybridised by connecting it with neural networks over ambiguous data in order to be applied to domains which ILP cannot address, while providing data efficiency and generalisation beyond what neural networks on their own can achieve.
Recommendations
- A comparative study of three neural-symbolic approaches to inductive logic programming
- First-Order Logical Neural Networks
- Neural probabilistic logic programming in DeepProbLog
- Generating rules with predicates, terms and variables from the pruned neural networks
- Artificial nonmonotonic neural networks
Cited in
(38)- Inductive logic programming at 30
- scientific article; zbMATH DE number 1706377 (Why is no real title available?)
- Making sense of sensory input
- Learning answer set programs with aggregates via sampling and genetic programming
- A differentiable first-order rule learner for inductive logic programming
- Injecting domain knowledge in neural networks: a controlled experiment on a constrained problem
- Synergies between machine learning and reasoning -- an introduction by the Kay R. Amel group
- Inductive learning of robot task knowledge from raw data and online expert feedback
- Towards unbiased exploration in partial label learning
- Semiring programming: a semantic framework for generalized sum product problems
- \(\alpha\)ILP: thinking visual scenes as differentiable logic programs
- Learning programs with magic values
- Learning Optimal Decision Sets and Lists with SAT
- Learning from interpretation transition using differentiable logic programming semantics
- Making sense of raw input
- Learning differentiable logic programs for abstract visual reasoning
- Learning logic programs by explaining their failures
- Rule Induction and Reasoning over Knowledge Graphs
- NeST: the neuro-symbolic transpiler
- A semantic framework for neurosymbolic computation
- Learning guided automated reasoning: a brief survey
- Neural logic rule layers
- Logic tensor networks
- Learning programs by learning from failures
- Top program construction and reduction for polynomial time meta-interpretive learning
- Meta-interpretive learning as metarule specialisation
- Explainable and interpretable machine learning and data mining
- Inductive general game playing
- Neural probabilistic logic programming in DeepProbLog
- Reinforcement learning meets logic programming: towards explainable AI
- Boolean matrix logic programming for active learning of gene functions in genome-scale metabolic network models
- FFNSL: Feed-forward neural-symbolic learner
- From statistical relational to neurosymbolic artificial intelligence: a survey
- Logical reduction of metarules
- Beyond graph neural networks with lifted relational neural networks
- A comparative study of three neural-symbolic approaches to inductive logic programming
- Learning temporal causal sequence relationships from real-time time-series
- Analyzing differentiable fuzzy logic operators
This page was built for publication: Learning explanatory rules from noisy data
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4605199)
