Standard
Fast automatic heuristic construction using active learning. / Ogilvie, William; Petoumenos, Pavlos
; Wang, Zheng et al.
Languages and Compilers for Parallel Computing: 27th International Workshop, LCPC 2014, Hillsboro, OR, USA, September 15-17, 2014, Revised Selected Papers. ed. / James Brodman; Peng Tu. Springer, 2014. p. 146-160 (Lecture Notes in Computer Science; Vol. 8967).
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
Harvard
Ogilvie, W, Petoumenos, P
, Wang, Z & Leather, H 2014,
Fast automatic heuristic construction using active learning. in J Brodman & P Tu (eds),
Languages and Compilers for Parallel Computing: 27th International Workshop, LCPC 2014, Hillsboro, OR, USA, September 15-17, 2014, Revised Selected Papers. Lecture Notes in Computer Science, vol. 8967, Springer, pp. 146-160.
https://doi.org/10.1007/978-3-319-17473-0_10
APA
Ogilvie, W., Petoumenos, P.
, Wang, Z., & Leather, H. (2014).
Fast automatic heuristic construction using active learning. In J. Brodman, & P. Tu (Eds.),
Languages and Compilers for Parallel Computing: 27th International Workshop, LCPC 2014, Hillsboro, OR, USA, September 15-17, 2014, Revised Selected Papers (pp. 146-160). (Lecture Notes in Computer Science; Vol. 8967). Springer.
https://doi.org/10.1007/978-3-319-17473-0_10
Vancouver
Ogilvie W, Petoumenos P
, Wang Z, Leather H.
Fast automatic heuristic construction using active learning. In Brodman J, Tu P, editors, Languages and Compilers for Parallel Computing: 27th International Workshop, LCPC 2014, Hillsboro, OR, USA, September 15-17, 2014, Revised Selected Papers. Springer. 2014. p. 146-160. (Lecture Notes in Computer Science). doi: 10.1007/978-3-319-17473-0_10
Author
Ogilvie, William ; Petoumenos, Pavlos
; Wang, Zheng et al. /
Fast automatic heuristic construction using active learning. Languages and Compilers for Parallel Computing: 27th International Workshop, LCPC 2014, Hillsboro, OR, USA, September 15-17, 2014, Revised Selected Papers. editor / James Brodman ; Peng Tu. Springer, 2014. pp. 146-160 (Lecture Notes in Computer Science).
Bibtex
@inproceedings{c471a35e17f944eb82eb8e10743b4757,
title = "Fast automatic heuristic construction using active learning",
abstract = "Building effective optimization heuristics is a challenging task which often takes developers several months if not years to complete. Predictive modelling has recently emerged as a promising solution, automatically constructing heuristics from training data. However, obtaining this data can take months per platform. This is becoming an ever more critical problem and if no solution is found we shall be left with out of date heuristics which cannot extract the best performance from modern machines.In this work, we present a low-cost predictive modelling approach for automatic heuristic construction which significantly reduces this training overhead. Typically in supervised learning the training instances are randomly selected to evaluate regardless of how much useful information they carry. This wastes effort on parts of the space that contribute little to the quality of the produced heuristic. Our approach, on the other hand, uses active learning to select and only focus on the most useful training examples.We demonstrate this technique by automatically constructing a model to determine on which device to execute four parallel programs at differing problem dimensions for a representative Cpu–Gpu based heterogeneous system. Our methodology is remarkably simple and yet effective, making it a strong candidate for wide adoption. At high levels of classification accuracy the average learning speed-up is 3x, as compared to the state-of-the-art.",
keywords = "Machine learning, Workload scheduling",
author = "William Ogilvie and Pavlos Petoumenos and Zheng Wang and Hugh Leather",
year = "2014",
month = sep,
doi = "10.1007/978-3-319-17473-0_10",
language = "English",
isbn = "9783319174723",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "146--160",
editor = "James Brodman and Peng Tu",
booktitle = "Languages and Compilers for Parallel Computing",
}
RIS
TY - GEN
T1 - Fast automatic heuristic construction using active learning
AU - Ogilvie, William
AU - Petoumenos, Pavlos
AU - Wang, Zheng
AU - Leather, Hugh
PY - 2014/9
Y1 - 2014/9
N2 - Building effective optimization heuristics is a challenging task which often takes developers several months if not years to complete. Predictive modelling has recently emerged as a promising solution, automatically constructing heuristics from training data. However, obtaining this data can take months per platform. This is becoming an ever more critical problem and if no solution is found we shall be left with out of date heuristics which cannot extract the best performance from modern machines.In this work, we present a low-cost predictive modelling approach for automatic heuristic construction which significantly reduces this training overhead. Typically in supervised learning the training instances are randomly selected to evaluate regardless of how much useful information they carry. This wastes effort on parts of the space that contribute little to the quality of the produced heuristic. Our approach, on the other hand, uses active learning to select and only focus on the most useful training examples.We demonstrate this technique by automatically constructing a model to determine on which device to execute four parallel programs at differing problem dimensions for a representative Cpu–Gpu based heterogeneous system. Our methodology is remarkably simple and yet effective, making it a strong candidate for wide adoption. At high levels of classification accuracy the average learning speed-up is 3x, as compared to the state-of-the-art.
AB - Building effective optimization heuristics is a challenging task which often takes developers several months if not years to complete. Predictive modelling has recently emerged as a promising solution, automatically constructing heuristics from training data. However, obtaining this data can take months per platform. This is becoming an ever more critical problem and if no solution is found we shall be left with out of date heuristics which cannot extract the best performance from modern machines.In this work, we present a low-cost predictive modelling approach for automatic heuristic construction which significantly reduces this training overhead. Typically in supervised learning the training instances are randomly selected to evaluate regardless of how much useful information they carry. This wastes effort on parts of the space that contribute little to the quality of the produced heuristic. Our approach, on the other hand, uses active learning to select and only focus on the most useful training examples.We demonstrate this technique by automatically constructing a model to determine on which device to execute four parallel programs at differing problem dimensions for a representative Cpu–Gpu based heterogeneous system. Our methodology is remarkably simple and yet effective, making it a strong candidate for wide adoption. At high levels of classification accuracy the average learning speed-up is 3x, as compared to the state-of-the-art.
KW - Machine learning
KW - Workload scheduling
U2 - 10.1007/978-3-319-17473-0_10
DO - 10.1007/978-3-319-17473-0_10
M3 - Conference contribution/Paper
SN - 9783319174723
T3 - Lecture Notes in Computer Science
SP - 146
EP - 160
BT - Languages and Compilers for Parallel Computing
A2 - Brodman, James
A2 - Tu, Peng
PB - Springer
ER -