A Proactive Sampling Approach to Project Scheduling under Uncertainty
Authors: Pradeep Varakantham, Na Fu, Hoong Chuin Lau
AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We provide an extensive experimental evaluation on benchmark problems from literature and compare against the best known approach to solve RCPSP/max with uncertainty. To demonstrate the utility of our proposed approaches, we compare against the best known work in the literature on generating α-robust makespan for RCPSP/max with durational uncertainty, referred to as FLPV (Fu et al. 2012). The problem instances considered in our experiments are obtained by extending the benchmark sets J10, J20 and J30 for RCPSP/max, as specified in PSPLib (Kolisch and Sprecher 1996). |
| Researcher Affiliation | Academia | Pradeep Varakantham, Na Fu and Hoong Chuin Lau School of Information Systems, Singapore Management University 80 Stamford Road, Singapore 178902 {pradeepv, nafu, hclau}@smu.edu.sg |
| Pseudocode | No | The paper presents mathematical optimization models (MILP formulations) in tables, but these are not pseudocode or algorithm blocks with procedural steps. |
| Open Source Code | No | The paper does not provide any explicit statement about releasing source code for the described methodology, nor does it include any links to a code repository. |
| Open Datasets | Yes | The problem instances considered in our experiments are obtained by extending the benchmark sets J10, J20 and J30 for RCPSP/max, as specified in PSPLib (Kolisch and Sprecher 1996). |
| Dataset Splits | No | The paper uses standard benchmark sets (J10, J20, J30) for evaluation, and while it performs preliminary experiments to tune parameters like γ and the number of samples, it does not describe explicit train/validation/test dataset splits for the problem instances themselves. |
| Hardware Specification | No | The paper describes runtimes of the algorithms but does not provide any specific hardware details such as CPU models, GPU types, or memory specifications used for the experiments. |
| Software Dependencies | No | The paper mentions exploiting 'commercial optimization software (e.g., CPLEX)' but does not provide a specific version number for CPLEX or any other software dependencies. |
| Experiment Setup | Yes | Unless otherwise specified, the default parameter values are σ = 0.5, α = 0.2, γ = 0.1 and Q = 35. |