On the Computation of Necessary and Sufficient Explanations

Authors: Adnan Darwiche, Chunxi Ji5582-5591

AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Table 1: Evaluating Algorithms 1 & 2. Times in secs. First ten entries are decision trees. Last five entries are decision graphs.
Researcher Affiliation Academia Computer Science Department University of California, Los Angeles darwiche@cs.ucla.edu, jich@cs.ucla.edu
Pseudocode Yes Algorithm 1: Shortest Necessary Reasons (SNRs); Algorithm 2: Shortest Sufficient Reasons (SSRs)
Open Source Code No The paper does not provide a specific link or an explicit statement about the public release of its source code.
Open Datasets Yes Table 1 depicts an empirical evaluation on decision trees learned from Open ML datasets (Vanschoren et al. 2013)...
Dataset Splits No The paper states, 'Each dataset was split using WEKA into training (85%) and testing (15%) data,' but it does not mention a validation split.
Hardware Specification Yes We used a Python implementation on a dual Intel(R) Xeon E5-2670 CPUs running at 2.60GHz and 256GB RAM.
Software Dependencies No The paper mentions using 'WEKA s J48 classifier' and 'python-weka-wrapper3 available at pypi.org' and a 'Python implementation' but does not specify exact version numbers for these software dependencies.
Experiment Setup No The paper mentions using 'WEKA s J48 classifier with default settings' for learning decision trees, but it does not provide specific hyperparameters or detailed training configurations for its own proposed algorithms (SNR and SSR).