Anytime Inference in Probabilistic Logic Programs with Tp-Compilation
Authors: Jonas Vlasselaer, Guy Van den Broeck, Angelika Kimmig, Wannes Meert, Luc De Raedt
IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | An empirical evaluation in these domains demonstrates that TP-compilation outperforms state-of-the-art sequential approaches on these problems with respect to time, space and quality of results. The paper is organized as follows. We review the necessary background in Section 2. Section 3 and 4 formally introduce the Tc P operator and corresponding algorithms. We discuss experimental results in Section 5 and conclude in Section 6. and 5 Experimental Results Our experiments address the following questions: Q1 How does TP-compilation compare to exact sequential WMC approaches? Q2 When is online TP-compilation advantageous? Q3 How does TP-compilation compare to anytime sequential approaches? Q4 What is the impact of approximating the model? |
| Researcher Affiliation | Academia | Jonas Vlasselaer, Guy Van den Broeck, Angelika Kimmig, Wannes Meert, Luc De Raedt Department of Computer Science KU Leuven, Belgium firstname.lastname@cs.kuleuven.be |
| Pseudocode | No | The paper describes algorithms and operators in prose (e.g., in Section 3 and 4) but does not contain a clearly labeled "Pseudocode" or "Algorithm" block or figure. |
| Open Source Code | Yes | Our implementation is available as part of the Prob Log package 4. http://dtai.cs.kuleuven.be/problog/ |
| Open Datasets | Yes | Smokers. Following Fierens et al. [2013], we generate random power law graphs for the standard Smokers social network domain. Alzheimer. We use series of connected subgraphs of the Alzheimer network of De Raedt et al. [2007]... Genes. Following Renkens et al. [2012; 2014], we use the biological network of Ourfali et al. [2007]... Web KB. We use the Web KB6 dataset restricted to the 100 most frequent words [Davis and Domingos, 2009]... 6http://www.cs.cmu.edu/webkb/ |
| Dataset Splits | No | The paper mentions training time and tests conducted, but it does not specify explicit dataset splits (e.g., percentages or counts) for training, validation, or testing data, nor does it refer to standard predefined splits for these purposes for the datasets used. |
| Hardware Specification | Yes | Experiments are run on a 3.4 GHz machine with 16 GB of memory. |
| Software Dependencies | No | The paper mentions software like the 'SDD package developed at UCLA', 'Prob Log package', 'c2d5', and 'Alchemy package', but it does not provide specific version numbers for any of these software dependencies. |
| Experiment Setup | Yes | The time budget is 5 minutes for Papr and 15 minutes for Porg (excluding the time to construct the relevant ground program). For MC-SAT, we sample either 5,000 or 10,000 times per variable in the CNF, which yields approximately the same runtime as our approach. and We compute relevant ground programs as well as CNFs (where applicable) following Fierens et al. [2013] and use the SDD package developed at UCLA3. Experiments are run on a 3.4 GHz machine with 16 GB of memory. Our implementation is available as part of the Prob Log package 4. and The selection procedure we employ picks the atom which maximizes the following heuristic value: WMC(Λi a) WMC(Λi 1 a ) φa (SIZE(Λia) SIZE(Λi 1 a ))/SIZE(Λi 1 a ) where SIZE(Λ) denotes the number of edges in SDD Λ and φa adjusts for the importance of a in proving queries. |