Algorithms and Applications for the Same-Decision Probability

Authors: S. J. Chen, A. Choi, A. Darwiche

JAIR 2014 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We propose the first exact algorithm for computing the SDP, and demonstrate its effectiveness on several real and synthetic networks. Finally, we present new complexity results, such as the complexity of computing the SDP on models with a Naive Bayes structure.
Researcher Affiliation Academia Suming Chen EMAIL Arthur Choi EMAIL Adnan Darwiche EMAIL Computer Science Department University of California, Los Angeles Los Angeles, CA 90095
Pseudocode Yes Algorithm 1 Computing the SDP in a Naive Bayes network.
Open Source Code No The paper does not provide concrete access to source code for the methodology described. It mentions a previous approximation algorithm by Choi et al. (2012) and third-party tools like BNGenerator, but no code release for their exact algorithm.
Open Datasets Yes Real networks were either learned from datasets provided by the UCI Machine Learning Repository (Bache & Lichman, 2013) or provided by HRL Laboratories and CRESST.
Dataset Splits No The paper mentions generating synthetic networks and selecting variables, e.g., 'randomly selected the decision variable, 25 query variables, and evidence variables.' For real networks, it uses 'at least 80% of the total network variables to be query variables.' However, it does not specify any training, validation, or test dataset splits in terms of percentages, counts, or predefined standard splits for reproducibility.
Hardware Specification No The paper describes running computations within specific time limits ('20 minute time limit', '2 hours') but does not specify any hardware details such as GPU/CPU models, memory, or other computer specifications used for the experiments.
Software Dependencies No The paper mentions using 'BNGenerator (Ide, Cozman, & Ramos, 2004)' for synthetic networks but does not provide specific version numbers for this or any other software, libraries, or programming languages used in the implementation.
Experiment Setup Yes For each network we selected at least 80% of the total network variables to be query variables so that we could emphasize how the size of the query set greatly influences the computation time. Each computation was given 20 minutes to complete. As we believe that the value of the threshold can greatly affect running time, we computed the SDP with thresholds T = [0.01, 0.1, 0.2, . . . , 0.8, 0.9, 0.99] and took the worst-case time.