We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Neural Datalog Through Time: Informed Temporal Modeling via Logical Specification

Abstract: Learning how to predict future events from patterns of past events is difficult when the set of possible event types is large. Training an unrestricted neural model might overfit to spurious patterns. To exploit domain-specific knowledge of how past events might affect an event's present probability, we propose using a temporal deductive database to track structured facts over time. Rules serve to prove facts from other facts and from past events. Each fact has a time-varying state---a vector computed by a neural net whose topology is determined by the fact's provenance, including its experience of past events. The possible event types at any time are given by special facts, whose probabilities are neurally modeled alongside their states. In both synthetic and real-world domains, we show that neural probabilistic models derived from concise Datalog programs improve prediction by encoding appropriate domain knowledge in their architecture.
Comments: ICML 2020 camera-ready (new Appendix A.3, rewritten Appendix F)
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Databases (cs.DB); Logic in Computer Science (cs.LO); Machine Learning (stat.ML)
Cite as: arXiv:2006.16723 [cs.LG]
  (or arXiv:2006.16723v2 [cs.LG] for this version)

Submission history

From: Hongyuan Mei [view email]
[v1] Tue, 30 Jun 2020 12:26:04 GMT (8482kb,D)
[v2] Mon, 17 Aug 2020 02:58:01 GMT (8503kb,D)

Link back to: arXiv, form interface, contact.