Exploit ILP to learn symmetry breaking constraints of ASP programs.

Overview

ILP Symmetry Breaking

Overview

This project aims to exploit inductive logic programming to lift symmetry breaking constraints of ASP programs.

Given an ASP file, we use the system SBASS (symmetry-breaking answer set solving) to infer its graph representation and then detect the symmetries as a graph automorphism problem (performed by the system SAUCY). SBASS returns a set of (irredundant) graph symmetry generators, which are used in our framework to compute the positive and negative examples for the ILP system ILASP.

Note: the files of Active Background Knowledge (active_BK/active_BK_sat) contain the constraints learned for the experiments. To test the framework, remove the constraints and follow the files' instructions to obtain the same result.

Project Structure

.
├── \Experiments              # Directory with experiments results 
│   ├── experiments.csv         # CSV file with results
│   └── experiments             # Script to compare the running-time performance     
│
├── \Instances              # Directory with problem instances
│   ├── \House_Configuration     # House-Configuration Problem     
│   ├── \Pigeon_Owner            # Pigeon-Hole Problem with colors and owners extension   
│   ├── \Pigeon_Color            # Pigeon-Hole Problem with colors extension
│   └── \Pigeon_Hole             # Pigeon-Hole Problem  
│
├── \src                    # Sources  
│   ├── \ILASP4                  # ILASP4 
│   ├── \SBASS                   # SBASS 
│   ├── file_names.py            # Python module with file names
│   ├── parser.py                # Main python file: create the positive and negative examples from SBASS output
│   ├── remove.py                # Auxiliary python file to remove duplicate in smodels file
│   └── permutations.lp          # ASP file which computes the (partial) non symmetric 
│                                  permutations of atoms
│
├── .gitignore 
├── .gitattributes
├── ILP_SBC                 # Script that runs SBASS and lift the SBC found using ILASP
└── README.md

Prerequisites

Usage

1) Create default positive examples

Create the default positive examples for Pigeon_Hole problem: each instance in the directory Gen generate a positive example.

$ .\ILP_SBC -g .\Instances\Pigeon_Hole

2) Create positive and negative examples

Default mode: each non-symmetric answer set defines a positive example

 $ .\ILP_SBC -d .\Instances\Pigeon_Hole

Satisfiable mode: define a single positive example with empty inclusions and exclusions

 $ .\ILP_SBC -s .\Instances\Pigeon_Hole

3) Run ILASP to extend the active background knowledge

 $ .\ILP_SBC -i .\Instances\Pigeon_Hole

Citations

C. Drescher, O. Tifrea, and T. Walsh, “Symmetry-breaking answer set solving” (SBASS)

@article{drescherSymmetrybreakingAnswerSet2011,
	title = {Symmetry-breaking answer set solving},
	volume = {24},
	doi = {10.3233/AIC-2011-0495},
	number = {2},
	journal = {AI Commun.},
	author = {Drescher, Christian and Tifrea, Oana and Walsh, Toby},
	year = {2011},
	pages = {177--194}
}

M. Law, A. Russo, and K. Broda, “The {ILASP} System for Inductive Learning of Answer Set Programs” (ILASP)

@article{larubr20b,
     title = {The {ILASP} System for Inductive Learning of Answer Set Programs},
     author = {M. Law and A. Russo  and K. Broda},
     journal = {The Association for Logic Programming Newsletter},
     year = {2020}
}
@misc{ilasp,
     author = {M. Law and A. Russo  and K. Broda},
     title = {Ilasp Releases},
     howpublished = {\url{www.ilasp.com}},
     note = {Accessed: 2020-10-01},
     year={2020}
}

This project uses reinforcement learning on stock market and agent tries to learn trading. The goal is to check if the agent can learn to read tape. The project is dedicated to hero in life great Jesse Livermore.

Reinforcement-trading This project uses Reinforcement learning on stock market and agent tries to learn trading. The goal is to check if the agent can

1.4k Dec 22, 2022

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

21.3k Jan 1, 2023

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

11.4k Feb 13, 2021

A static analysis library for computing graph representations of Python programs suitable for use with graph neural networks.

python_graphs This package is for computing graph representations of Python programs for machine learning applications. It includes the following modu

258 Dec 29, 2022

Code for ICML 2021 paper: How could Neural Networks understand Programs?

Exploit ILP to learn symmetry breaking constraints of ASP programs.

Related tags

Overview

ILP Symmetry Breaking

Overview

Project Structure

Prerequisites

Usage

1) Create default positive examples

2) Create positive and negative examples

Default mode: each non-symmetric answer set defines a positive example

Satisfiable mode: define a single positive example with empty inclusions and exclusions

3) Run ILASP to extend the active background knowledge

Citations

You might also like...

This project uses reinforcement learning on stock market and agent tries to learn trading. The goal is to check if the agent can learn to read tape. The project is dedicated to hero in life great Jesse Livermore.

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

A static analysis library for computing graph representations of Python programs suitable for use with graph neural networks.

Code for ICML 2021 paper: How could Neural Networks understand Programs?

[ICML 2021] Break-It-Fix-It: Learning to Repair Programs from Unlabeled Data

This repository contains the code for the paper "Hierarchical Motion Understanding via Motion Programs"

TensorFlowOnSpark brings TensorFlow programs to Apache Spark clusters.

PerfFuzz: Automatically Generate Pathological Inputs for C/C++ programs

Owner

Research Group Production Systems

This repo is a C++ version of yolov5_deepsort_tensorrt. Packing all C++ programs into .so files, using Python script to call C++ programs further.

Breaking the Dilemma of Medical Image-to-image Translation

Official implementation for "Low-light Image Enhancement via Breaking Down the Darkness"

DC3: A Learning Method for Optimization with Hard Constraints

Source code for the Paper: CombOptNet: Fit the Right NP-Hard Problem by Learning Integer Programming Constraints}

[WACV 2020] Reducing Footskate in Human Motion Reconstruction with Ground Contact Constraints

PyTorch reimplementation of hand-biomechanical-constraints (ECCV2020)

Angora is a mutation-based fuzzer. The main goal of Angora is to increase branch coverage by solving path constraints without symbolic execution.

Paper: Cross-View Kernel Similarity Metric Learning Using Pairwise Constraints for Person Re-identification

FindFunc is an IDA PRO plugin to find code functions that contain a certain assembly or byte pattern, reference a certain name or string, or conform to various other constraints.