Andrew Cropper
I run the logic and learning group at the University of Oxford. We work on logic and machine learning (inductive logic programming) (ILP). We are building an Automatic Computer Scientist. We develop the ILP system Popper.
News
Aug 24 Josh Rule's paper on explaining human learning through symbolic learning has been accepted for Nature Comms.
Jul 24 One paper accepted for ECAI24:
Learning logic programs by finding minimal unsatisfiable subprograms (with C. Hocquette)
Apr 24 Two papers accepted for IJCAI24:
Learning big logical rules by joining small rules (with C. Hocquette, A. Niskanen, R. Morel, and M. Järvisalo)
Learning logic programs by discovering higher-order abstractions (with C. Hocquette and S. Dumančić)
Feb 24 Céline gave a talk on constraint programming for inductive logic programming at AAAI24
Dec 23 Two papers accepted for AAAI24:
Learning MDL logic programs from noisy data (with C. Hocquette, A. Niskanen, and M. Järvisalo)
Generalisation through negation and predicate invention (with D. Cerna)
Nov 23 Rolf passed his DPhil/PhD viva!
Oct 23 Minghao Liu joined our group as a postdoc.
Oct 23 Céline presented our paper learning logic programs by combing programs (slides) at ECAI23.
Sep 23 Filipe Gouveia joined our group as a postdoc.
May 23 Rolf's paper on explaining the failures of logic programs has been accepted for the MLJ.
Feb 23 We gave a tutorial on inductive logic programming at AAAI23.
Feb 23 I gave a new faculty talk at AAAI23.
Feb 23 We presented our MLJ paper learning programs by learning from failures (slides) in the journal track at AAAI23.
Feb 23 We presented two papers at AAAI23:
Relational program synthesis with numerical reasoning (with C. Hocquette)
Learning logic programs by discovering where not to search (with C. Hocquette)