DeepStochLog: Neural Stochastic Logic Programming
This addresses computational efficiency problems for researchers in neural symbolic learning, though it appears incremental as an alternative framework building on existing stochastic logic programs.
The authors tackled the computational hardness of inference in neural probabilistic logic programs by proposing DeepStochLog, a neural symbolic framework based on stochastic definite clause grammars that defines probability distributions over derivations. They showed that DeepStochLog scales much better for inference and learning than existing approaches and achieves state-of-the-art results on challenging tasks.
Recent advances in neural symbolic learning, such as DeepProbLog, extend probabilistic logic programs with neural predicates. Like graphical models, these probabilistic logic programs define a probability distribution over possible worlds, for which inference is computationally hard. We propose DeepStochLog, an alternative neural symbolic framework based on stochastic definite clause grammars, a type of stochastic logic program, which defines a probability distribution over possible derivations. More specifically, we introduce neural grammar rules into stochastic definite clause grammars to create a framework that can be trained end-to-end. We show that inference and learning in neural stochastic logic programming scale much better than for neural probabilistic logic programs. Furthermore, the experimental evaluation shows that DeepStochLog achieves state-of-the-art results on challenging neural symbolic learning tasks.