Hyperparameter Tuning in Bandit-Based Adaptive Operator Selection
Name
Amarasinghe-Hyperparameter Tuning.pdf
Size
359.74 KB
Format
Adobe PDF
Checksum (MD5)
702bd9d078e093e8384ae6b18515feee
Author(s)
Pacula, Maciej
Ansel, Jason Andrew
Amarasinghe, Saman P.
O'Reilly, Una-May
Date Issued
March 2012
Journal
Applications of Evolutionary Computation
Publisher
Springer Berlin / Heidelberg
Citation
Pacula, Maciej et al. “Hyperparameter Tuning in Bandit-Based Adaptive Operator Selection.” Applications of Evolutionary Computation. Ed. Cecilia Chio et al. LNCS Vol. 7248. Berlin, Heidelberg: Springer Berlin Heidelberg, 2012. 73–82.
Version
Author's final manuscript
Abstract
We are using bandit-based adaptive operator selection while autotuning parallel computer programs. The autotuning, which uses evolutionary algorithm-based stochastic sampling, takes place over an extended duration and occurs in situ as programs execute. The environment or context during tuning is either largely static in one scenario or dynamic in another. We rely upon adaptive operator selection to dynamically generate worthy test configurations of the program. In this paper, we study how the choice of hyperparameters, which control the trade-off between exploration and exploitation, affects the effectiveness of adaptive operator selection which in turn affects the performance of the autotuner. We show that while the optimal assignment of hyperparameters varies greatly between different benchmarks, there exists a single assignment, for a context, of hyperparameters that performs well regardless of the program being tuned.
Description
EvoApplications 2012: EvoCOMNET, EvoCOMPLEX, EvoFIN, EvoGAMES, EvoHOT, EvoIASP, EvoNUM, EvoPAR, EvoRISK, EvoSTIM, and EvoSTOC, Málaga, Spain, April 11-13, 2012, Proceedings
MIT Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike 3.0
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1007/978-3-642-29178-4_8
