Shortest Path Based Decision Making Using Probabilistic Inference
Authors: Akshat Kumar
AAAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We present results comparing EM against the MIP solver Cplex v12.6 for the RNDP problem. Figures 3 and 4 show the quality comparisons between EM and Cplex for a range of budgets. |
| Researcher Affiliation | Academia | Akshat Kumar School of Information Systems Singapore Management University akshatkumar@smu.edu.sg |
| Pseudocode | Yes | Algorithm 1: path EM(G = (V, E), s, t) |
| Open Source Code | No | The paper does not provide concrete access to source code (no repository link, explicit code release statement, or mention of code in supplementary materials). |
| Open Datasets | No | We used grid shaped graphs to simulate realistic road networks, with sizes ranging from 5x5 grid to 20x20 grid. The paper describes how these graphs were generated and their properties, but does not refer to a publicly available dataset with a specific source or citation. |
| Dataset Splits | No | The paper does not specify exact split percentages or sample counts for training, validation, or test sets. It describes the generation of grid graphs but not their partitioning for evaluation. |
| Hardware Specification | Yes | All our experiments are performed on a 16 core linux machine (with 32 parallel threads). |
| Software Dependencies | Yes | We present results comparing EM against the MIP solver Cplex v12.6 for the RNDP problem. |
| Experiment Setup | Yes | Both EM and Cplex were allowed to use 20 parallel threads with 10GB RAM limit. The time cutoff was 3 hours for each algorithm per instance. Each data point is an average over 5 randomly generated instances. We used a fixed penalty weight ρ = 0.005 for each edge and started applying penalty from iteration 1000 onwards. |