Multi-Objective Influence Diagrams with Possibly Optimal Policies

Authors: Radu Marinescu, Abdul Razak, Nic Wilson

AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We compare variants of the algorithm experimentally.
Researcher Affiliation Collaboration Radu Marinescu IBM Research Ireland radu.marinescu@ie.ibm.com Abdul Razak and Nic Wilson Insight Centre for Data Analytics University College Cork, Ireland {abdul.razak,nic.wilson}@insight-centre.org
Pseudocode Yes Algorithm 1: MOVE(OPT)
Open Source Code No The paper does not contain any explicit statement or link indicating that the source code for the methodology is openly available.
Open Datasets No The problem instances were generated using the random problem generator from (Marinescu, Razak, and Wilson 2012) using 2, 3 and 5 objectives, respectively. (No specific public access details for the generated instances or the generator were provided for reproducibility.)
Dataset Splits No The paper uses generated problem instances but does not specify any training/validation/test splits.
Hardware Specification Yes All algorithms were written in C++ and the experiments were run on a 2.6GHz processor with 4GB of RAM.
Software Dependencies No The paper states 'All algorithms were written in C++', but does not provide specific version numbers for compilers, libraries, or other software dependencies.
Experiment Setup Yes All algorithms were allotted a 20 minute time limit per problem instance. For each problem size (characterized by the number of chance variables C, number of decision variables D, number of objectives O), we generated 100 random instances. The random tradeoffs are mainly characterized by the parameters K and T, where K is the number of pairwise or binary tradeoffs and T is the number of 3-way tradeoffs, respectively.