Robust Decision Making for Stochastic Network Design

Authors: Akshat Kumar, Arambam Singh, Pradeep Varakantham, Daniel Sheldon

AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Empirically, we show that previous approaches that work on point estimates of network parameters result in high regret on several standard benchmarks, while our approach provides significantly more robust solutions.
Researcher Affiliation Academia School of Information Systems, Singapore Management University {akshatkumar,ajsingh,pradeepv}@smu.edu.sg College of Information and Computer Sciences, University of Massachusetts Amherst tsheldon@cs.umass.edu
Pseudocode No The paper includes 'Table 1: Constraint set Ω for max-regret mixed-integer program', which lists mathematical constraints, but not a pseudocode block or a clearly labeled algorithm block.
Open Source Code No The paper does not provide any explicit statement or link indicating that the source code for the methodology is openly available.
Open Datasets Yes We used a publicly available conservation planning benchmark which represents a geographical region on the coast of North Carolina (Ahmadizadeh et al. 2010).
Dataset Splits No The paper mentions 'N training cascades' for the SAA approximation but does not explicitly provide information on training/validation/test dataset splits, percentages, or sample counts for reproducibility.
Hardware Specification No The paper does not provide specific hardware details such as GPU or CPU models, processor types, or memory specifications used for running the experiments.
Software Dependencies No The paper mentions using a 'mixed-integer solver' and 'Python', but does not provide specific version numbers for any software components, libraries, or solvers.
Experiment Setup Yes Parameter Setting We first compute pbase e for each edge e using the equations provided in (Sheldon et al. 2010). Then we add uncertainty ϵ to each parameter to get the range [pbase e ϵ pbase e , pbase e +ϵ pbase e ]. The range for suitability scores is also computed in a similar manner. Edge Activation Regret Analysis For this analysis, we used a time horizon of T = 5, 9, 10, number of SAA samples N = 20, and the budget B = 10%. We set uncertainty level ϵ=30% as it provided pronounced effects of the MMR solution while also being a realistic level of noise.