Risk-Aware Proactive Scheduling via Conditional Value-at-Risk

Authors: Wen Song, Donghun Kang, Jie Zhang, Hui Xi

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Empirical results show that our algorithm outperforms state-of-the-art approaches with higher solution quality. We conduct extensive experiments on benchmark instances and commonly used uncertainty models.
Researcher Affiliation Collaboration Rolls-Royce@NTU Corp Lab, Nanyang Technological University, Singapore a School of Computer Science and Engineering, Nanyang Technological University, Singapore b Rolls-Royce Singapore Pte Ltd, Singapore
Pseudocode Yes Algorithm 1: Bn B(GV , G H, ω , ˆF β)
Open Source Code No The paper does not provide any statement about releasing source code for the described methodology, nor does it provide a link to a code repository.
Open Datasets No The paper generates its own RCPSP instances using 'Ran Gen2 (Vanhoucke et al. 2008)' and models uncertainty using 'a normal distribution N(d0 i , σ2)' and 'an exponential distribution Exp(1/d0 i )'. It does not refer to or provide access to a specific publicly available dataset used for training in the conventional sense.
Dataset Splits No The paper mentions using 'Q samples independently drawn from y' for SAA and 'Qt = 2000 testing samples' for PoF computation, but it does not specify explicit training/validation/test dataset splits with percentages, sample counts, or references to predefined splits as is common for reproducibility.
Hardware Specification Yes All algorithms run on an Intel Xeon E5 Workstation (3.5GHz, 16GB).
Software Dependencies Yes Our algorithm is implemented in JAVA 1.8, while SORU-H and BACCHUS are coded using Java API for CPLEX 12.7.1.
Experiment Setup Yes α is set to 0.2 in this section. We conduct experiments on the 72 instances used in Section 6.1, with ϵ {0, 0.05, 0.1}. In the remaining experiments, we set Q = 100.