Hybrid Models for Learning to Branch

Authors: Prateek Gupta, Maxime Gasse, Elias Khalil, Pawan Mudigonda, Andrea Lodi, Yoshua Bengio

NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate our methods on four classes of MILP problems, and show that they lead to up to 26% reduction in solver running time compared to state-of-the-art methods without a GPU, while extrapolating to harder problems than it was trained on.
Researcher Affiliation Academia Prateek Gupta University of Oxford The Alan Turing Institute pgupta@robots.ox.ac.uk Maxime Gasse Mila, Polytechnique Montréal maxime.gasse@polymol.ca Elias B. Khalil University of Toronto khalil@mie.utoronto.ca M. Pawan Kumar University of Oxford pawan@robots.ox.ac.uk Andrea Lodi CERC, Polytechnique Montréal andrea.lodi@polymol.ca Yoshua Bengio Mila, Université de Montréal yoshua.bengio@mila.quebec
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code Yes The code for this project is publicly available at https: //github.com/pg2455/Hybrid-learn2branch.
Open Datasets No The paper states 'Randomly generated instances are solved offline using SCIP [20] to collect training samples of the form (G0, G, X, i SB).' and 'We leave the description of the data collection and training details of each model to the supplementary materials.' It does not provide concrete access information (link, DOI, formal citation with authors/year) for a publicly available or open dataset.
Dataset Splits Yes To overcome this issue we used weight decay [34] regularization, with a validation set of 2000 observations generated using random medium instances (not used for evaluation).
Hardware Specification No The paper mentions running on 'GPU' and 'CPU' and refers to a 'high-end GPU card' but does not provide specific hardware models (e.g., NVIDIA A100, Intel Core i7) or detailed specifications used for the experiments.
Software Dependencies No The paper mentions the use of 'SCIP [20]' for solving instances and evaluation, but does not explicitly state the version number of SCIP or other software dependencies within the main text.
Experiment Setup Yes Each scenario uses 20 instances, solved using 3 different seeds to account for solver variability. All branching strategies are evaluated using the open-source solver SCIP [20] with a time limit of 45 minutes, and cutting planes are allowed only at the root node. To overcome this issue we used weight decay [34] regularization, with a validation set of 2000 observations generated using random medium instances (not used for evaluation).