DeepProbLog

DeepProbLog is an extension of ProbLog that integrates Probabilistic Logic Programming with deep learning by introducing the neural predicate. The neural predicate represents probabilistic facts whose probabilites are parameterized by neural networks. For more information, consult the papers listed below.

Installation

DeepProbLog can easily be installed using the following command: Make sure the following packages are installed:

pip install deepproblog

Test

To make sure your installation works, install pytest

pip install pytest

and run

python -m deepproblog test

Requirements

DeepProbLog has the following requirements:

• Python > 3.9
• ProbLog
• PySDD
• PyTorch
• TorchVision

Approximate Inference

To use Approximate Inference, we have the following additional requirements

• PySwip
  • Use pip install git+https://github.com/ML-KULeuven/pyswip
• SWI-Prolog < 9.0.0 The latter can be installed on Ubuntu with the following commands:

  sudo apt-add-repository ppa:swi-prolog/stable
  sudo apt install swi-prolog=8.4* swi-prolog-nox=8.4* swi-prolog-x=8.4*

  Experiments

The experiments are presented in the papers are available in the src/deepproblog/examples directory.

Papers

1. Robin Manhaeve, Sebastijan Dumancic, Angelika Kimmig, Thomas Demeester, Luc De Raedt: DeepProbLog: Neural Probabilistic Logic Programming. NeurIPS 2018: 3753-3763 (paper)
2. Robin Manhaeve, Sebastijan Dumancic, Angelika Kimmig, Thomas Demeester, Luc De Raedt: Neural Probabilistic Logic Programming in DeepProbLog. AIJ (paper)
3. Robin Manhaeve, Giuseppe Marra, Luc De Raedt: Approximate Inference for Neural Probabilistic Logic Programming. KR 2021

   License

   Copyright 2023 KU Leuven, DTAI Research Group

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.