Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
Deep Bucket Elimination
Authors: Yasaman Razeghi, Kalev Kask, Yadong Lu, Pierre Baldi, Sakshi Agarwal, Rina Dechter
IJCAI 2021 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our empirical results show that DBE is overall significantly more accurate than WMB especially on hard instances and even when the latter is provided the most generous memory resources feasible. |
| Researcher Affiliation | Academia | University of California, Irvine EMAIL |
| Pseudocode | Yes | Algorithm 1: [Deep] Bucket Elimination (DBE) Algorithm 2: approximate-NN(λ, ϵ) |
| Open Source Code | Yes | We provided the source code to reproduce the results of this paper at https://github.com/dechterlab/DBE. |
| Open Datasets | Yes | We carried our experiments on instances selected from three well-known benchmarks from the UAI repository used in [Kask et al., 2020] such as grids (vision domain), Pedigrees (from genetic linkage analysis), and DBNs. |
| Dataset Splits | Yes | Once the samples are available, we split them into training (80%), validation (10%), and test sets (10%). |
| Hardware Specification | No | The paper does not provide any specific hardware details such as CPU/GPU models, memory, or cloud instance types used for running the experiments. |
| Software Dependencies | No | The paper mentions the Adam optimizer and Sherpa software but does not provide specific version numbers for any software dependencies or libraries. |
| Experiment Setup | Yes | We train the network using the Adam optimizer [Kingma and Ba, 2014] with a learning rate of 0.001 and a batchsize of 256. In all the experiments, we used 5 105 samples for training the NNs with an error bound of ϵ = 10 6. The #epochs was bounded at 100. |