Meta-Level Control of Anytime Algorithms with Online Performance Prediction
Authors: Justin Svegliato, Kyle Hollins Wray, Shlomo Zilberstein
IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Most importantly, we show that our approach outperforms existing techniques that require substantial offline work on several benchmark domains and a mobile robot domain. |
| Researcher Affiliation | Academia | Justin Svegliato and Kyle Hollins Wray and Shlomo Zilberstein College of Information and Computer Sciences University of Massachusetts Amherst {jsvegliato,wray,shlomo}@cs.umass.edu |
| Pseudocode | Yes | Algorithm 1: A meta-level control technique that uses online performance prediction. ... Algorithm 1 describes our meta-level control technique. |
| Open Source Code | No | The paper does not provide an explicit statement or link for the open-source code of their own methodology. It only mentions using existing open-source implementations for benchmark algorithms. |
| Open Datasets | No | The paper describes various benchmark domains and a mobile robot domain but does not provide concrete access information (links, DOIs, or specific citations for the datasets used in their experiments) for these problem instances. |
| Dataset Splits | No | The paper does not provide specific dataset split information (percentages, sample counts, or citations to predefined splits) needed to reproduce data partitioning. |
| Hardware Specification | No | The paper mentions running experiments on 'an i Clebo Kobuki' mobile robot but does not provide specific computational hardware details (CPU/GPU models, memory, or cloud instance specifications) used for running its simulations or training algorithms. |
| Software Dependencies | No | The paper mentions several software components (e.g., SciPy, jsp-ga, QAPLIB, epic) but does not provide specific version numbers for any of them. |
| Experiment Setup | No | The paper describes the general experimental procedure and the functional form of the performance predictor, but does not provide specific hyperparameter values or detailed training configurations for the algorithms or the meta-level control technique. |