Code for "Unifying Local Consistency and MAX SAT Relaxations for Scalable Inference with Rounding Guarantees." Stephen H. Bach, Bert Huang, and Lise Getoor. Artificial Intelligence and Statistics (AISTATS) 2015
Groovy Java Python Shell
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
data
src/main
.gitignore
LICENSE
README.md
pom.xml
run.sh

README.md

bach-aistats15-code

Code for "Unifying Local Consistency and MAX SAT Relaxations for Scalable Inference with Rounding Guarantees." Stephen H. Bach, Bert Huang, and Lise Getoor. Artificial Intelligence and Statistics (AISTATS) 2015

Requirements

  • Maven 3 - Used to install PSL and manage dependencies.
  • MPLP - Dual decomposition implementation, see the MPLP README.
    • Compile and place a copy of solver on the executable path.
    • Edit line 27 of cycle_tighten_main.cpp to define MAX_TIGHT_ITERS as 0. Recompile and place of copy of solver on the executable path with the name solver-nocycles.

Running the experiments

Execute run.sh.