Current browse context:
stat.ML
Change to browse by:
References & Citations
Statistics > Machine Learning
Title: Theory reconstruction: a representation learning view on predicate invention
(Submitted on 28 Jun 2016 (v1), last revised 29 Jun 2016 (this version, v2))
Abstract: With this positional paper we present a representation learning view on predicate invention. The intention of this proposal is to bridge the relational and deep learning communities on the problem of predicate invention. We propose a theory reconstruction approach, a formalism that extends autoencoder approach to representation learning to the relational settings. Our intention is to start a discussion to define a unifying framework for predicate invention and theory revision.
Submission history
From: Sebastijan Dumancic [view email][v1] Tue, 28 Jun 2016 11:41:03 GMT (12kb)
[v2] Wed, 29 Jun 2016 11:29:35 GMT (12kb)
Link back to: arXiv, form interface, contact.