Machine Learning for Integer Programming
Authors: Elias B. Khalil
IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | To illustrate the potential for ML in MIP, I have so far tackled branching variable selection, a crucial component of the search procedure, showing that ML approaches for variable selection can outperform traditional heuristics. ... Our approach, SB+ML, significantly outperforms competing methods, solving more instances than both PC and SB+PC (a hybrid of SB and PC), while also requiring around 36% and 16% fewer nodes on average, respectively. In terms of running time, our method incurs some additional overhead pernode due to feature computations, yet it can be 15 to 20% faster than the competitors overall on instances of medium and hard difficulties, thanks to large savings in the number of nodes processed. |
| Researcher Affiliation | Academia | Elias B. Khalil School of Computational Science & Engineering Georgia Institute of Technology ekhalil3@gatech.edu |
| Pseudocode | No | The paper describes the approaches but does not include any pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any statement about releasing source code or a link to a code repository. |
| Open Datasets | No | The paper mentions collecting data for training ('the collected data being used for training') but does not provide concrete access information (e.g., link, citation to a specific dataset, or repository) for this data. It refers to a previous paper for experimental setup details. |
| Dataset Splits | No | The paper does not explicitly provide details about training, validation, or test splits. It states: 'We do not give details of the experimental setup here, and refer to our paper instead [Khalil et al., 2016].' |
| Hardware Specification | No | The paper does not provide any specific hardware details such as GPU/CPU models, processor types, or memory used for experiments. |
| Software Dependencies | No | The paper mentions commercial and non-commercial MIP solvers like IBM CPLEX and SCIP but does not specify their version numbers or any other software dependencies with version numbers used in their own experiments. |
| Experiment Setup | No | The paper explicitly states: 'We do not give details of the experimental setup here, and refer to our paper instead [Khalil et al., 2016].' Therefore, specific experimental setup details are not provided in this paper. |