New Rules for Domain Independent Lifted MAP Inference
Authors: Happy Mittal, Prasoon Goyal, Vibhav G Gogate, Parag Singla
NeurIPS 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We compared the performance of our algorithm against Sarkhel et. al [19] s non shared MLN approach and the purely grounded version on three benchmark MLNs. For both the lifted approaches, we used them as pre-processing algorithms to reduce the MLN domains. We applied the ILP based solver Gurobi [8] as the base solver on the reduced theory to find the MAP assignment. [...] For each algorithm, we report: 1) Time: Time to reach the optimal as the domain size is varied from 25 to 1000. 2) Cost: Cost of the unsatisfied clauses as the running time is varied for a fixed domain size (500). 3) Theory Size: Ground theory size as the domain size is varied. All the experiments were run on an Intel four core i3 processor with 4 GB of RAM. |
| Researcher Affiliation | Academia | Happy Mittal, Prasoon Goyal Dept. of Comp. Sci. & Engg. I.I.T. Delhi, Hauz Khas New Delhi, 110016, India [...] Vibhav Gogate Dept. of Comp. Sci. Univ. of Texas Dallas Richardson, TX 75080, USA [...] Parag Singla Dept. of Comp. Sci. & Engg. I.I.T. Delhi, Hauz Khas New Delhi, 110016, India |
| Pseudocode | Yes | Algorithm 1 Reducing all the single occurrence equivalence classes in an MLN [...] Algorithm 2 Finding Tautologies at Extremes with Single Occurrence Variables |
| Open Source Code | No | The paper does not provide any explicit statement or link indicating that its own source code is publicly available. |
| Open Datasets | Yes | We used the following benchmark MLNs for our experiments. [...] 1) Information Extraction (IE): This theory is available from the Alchemy [13] website. [...] 2) Friends & Smokers (FS): This is a standard MLN used earlier in the literature [20]. |
| Dataset Splits | No | The paper does not specify dataset splits (e.g., training, validation, test) for its experiments. It focuses on the time and cost of inference on theoretical MLN structures with varying domain sizes. |
| Hardware Specification | Yes | All the experiments were run on an Intel four core i3 processor with 4 GB of RAM. |
| Software Dependencies | Yes | We applied the ILP based solver Gurobi [8] as the base solver on the reduced theory to find the MAP assignment. [8] Gurobi Optimization Inc. Gurobi Optimizer Reference Manual, 2013. http://gurobi.com. |
| Experiment Setup | Yes | For each algorithm, we report: 1) Time: Time to reach the optimal as the domain size is varied from 25 to 1000. 2) Cost: Cost of the unsatisfied clauses as the running time is varied for a fixed domain size (500). 3) Theory Size: Ground theory size as the domain size is varied. [...] For IE, two of the variable domains of were varied and other two were kept constant at 10 as done in [19]. Reported results are averaged over 5 runs. |