
DeepLogic: EndtoEnd Logical Reasoning
Neural networks have been learning complex multihop reasoning in variou...
read it

A Probabilistic Extension of Action Language BC+
We present a probabilistic extension of action language BC+. Just like B...
read it

Probability Aggregates in Probability Answer Set Programming
Probability answer set programming is a declarative programming that has...
read it

Strong equivalence for LP^MLN programs
Strong equivalence is a wellstudied and important concept in answer set...
read it

NeuroSymbolic Constraint Programming for Structured Prediction
We propose Nester, a method for injecting neural networks into constrain...
read it

Neurocoder: Learning GeneralPurpose Computation Using Stored Neural Programs
Artificial Neural Networks are uniquely adroit at machine learning by pr...
read it

A Unified Framework for Nonmonotonic Reasoning with Vagueness and Uncertainty
An answer set programming paradigm is proposed that supports nonmonotoni...
read it
Extending Answer Set Programs with Neural Networks
The integration of lowlevel perception with highlevel reasoning is one of the oldest problems in Artificial Intelligence. Recently, several proposals were made to implement the reasoning process in complex neural network architectures. While these works aim at extending neural networks with the capability of reasoning, a natural question that we consider is: can we extend answer set programs with neural networks to allow complex and highlevel reasoning on neural network outputs? As a preliminary result, we propose NeurASP – a simple extension of answer set programs by embracing neural networks where neural network outputs are treated as probability distributions over atomic facts in answer set programs. We show that NeurASP can not only improve the perception accuracy of a pretrained neural network, but also help to train a neural network better by giving restrictions through logic rules. However, training with NeurASP would take much more time than pure neural network training due to the internal use of a symbolic reasoning engine. For future work, we plan to investigate the potential ways to solve the scalability issue of NeurASP. One potential way is to embed logic programs directly in neural networks. On this route, we plan to first design a SAT solver using neural networks, then extend such a solver to allow logic programs.
READ FULL TEXT
Comments
There are no comments yet.