Accuracy-aware optimization of approximate programs

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Many modern applications (such as multimedia processing, machine learning, and big-data analytics) exhibit an inherent tradeoff between performance and the accuracy of the produced results. These applications allow us to investigate new, more aggressive program optimizations. We present a novel approximate optimization framework based on accuracy-aware program transformations. These transformations reduce accuracy of computation in return for improved performance, energy efficiency, and/or resilience. The optimization framework includes program analyses that characterize the accuracy of transformed programs, and search techniques that navigate the tradeoff space induced by transformations to find approximate programs with profitable tradeoffs. Our research shows that this approach can significantly improve program performance while providing acceptably accurate results.

Original languageEnglish (US)
Title of host publication2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages37-38
Number of pages2
ISBN (Electronic)9781467383202
DOIs
StatePublished - Nov 10 2015
EventInternational Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015 - Amsterdam, Netherlands
Duration: Oct 4 2015Oct 9 2015

Publication series

Name2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015

Other

OtherInternational Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015
CountryNetherlands
CityAmsterdam
Period10/4/1510/9/15

Fingerprint

Energy efficiency
Learning systems
Processing
Big data

Keywords

  • Accuracy
  • Approximation methods
  • Computational efficiency
  • Hardware
  • Optimization
  • Probabilistic logic
  • Programming

ASJC Scopus subject areas

  • Hardware and Architecture
  • Computer Science Applications

Cite this

Misailovic, S. (2015). Accuracy-aware optimization of approximate programs. In 2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015 (pp. 37-38). [7324543] (2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CASES.2015.7324543

Accuracy-aware optimization of approximate programs. / Misailovic, Sasa.

2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015. Institute of Electrical and Electronics Engineers Inc., 2015. p. 37-38 7324543 (2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Misailovic, S 2015, Accuracy-aware optimization of approximate programs. in 2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015., 7324543, 2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015, Institute of Electrical and Electronics Engineers Inc., pp. 37-38, International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015, Amsterdam, Netherlands, 10/4/15. https://doi.org/10.1109/CASES.2015.7324543
Misailovic S. Accuracy-aware optimization of approximate programs. In 2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015. Institute of Electrical and Electronics Engineers Inc. 2015. p. 37-38. 7324543. (2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015). https://doi.org/10.1109/CASES.2015.7324543
Misailovic, Sasa. / Accuracy-aware optimization of approximate programs. 2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015. Institute of Electrical and Electronics Engineers Inc., 2015. pp. 37-38 (2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015).
@inproceedings{ba47acfde93a405db5305b3a3a5f5811,
title = "Accuracy-aware optimization of approximate programs",
abstract = "Many modern applications (such as multimedia processing, machine learning, and big-data analytics) exhibit an inherent tradeoff between performance and the accuracy of the produced results. These applications allow us to investigate new, more aggressive program optimizations. We present a novel approximate optimization framework based on accuracy-aware program transformations. These transformations reduce accuracy of computation in return for improved performance, energy efficiency, and/or resilience. The optimization framework includes program analyses that characterize the accuracy of transformed programs, and search techniques that navigate the tradeoff space induced by transformations to find approximate programs with profitable tradeoffs. Our research shows that this approach can significantly improve program performance while providing acceptably accurate results.",
keywords = "Accuracy, Approximation methods, Computational efficiency, Hardware, Optimization, Probabilistic logic, Programming",
author = "Sasa Misailovic",
year = "2015",
month = "11",
day = "10",
doi = "10.1109/CASES.2015.7324543",
language = "English (US)",
series = "2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "37--38",
booktitle = "2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015",
address = "United States",

}

TY - GEN

T1 - Accuracy-aware optimization of approximate programs

AU - Misailovic, Sasa

PY - 2015/11/10

Y1 - 2015/11/10

N2 - Many modern applications (such as multimedia processing, machine learning, and big-data analytics) exhibit an inherent tradeoff between performance and the accuracy of the produced results. These applications allow us to investigate new, more aggressive program optimizations. We present a novel approximate optimization framework based on accuracy-aware program transformations. These transformations reduce accuracy of computation in return for improved performance, energy efficiency, and/or resilience. The optimization framework includes program analyses that characterize the accuracy of transformed programs, and search techniques that navigate the tradeoff space induced by transformations to find approximate programs with profitable tradeoffs. Our research shows that this approach can significantly improve program performance while providing acceptably accurate results.

AB - Many modern applications (such as multimedia processing, machine learning, and big-data analytics) exhibit an inherent tradeoff between performance and the accuracy of the produced results. These applications allow us to investigate new, more aggressive program optimizations. We present a novel approximate optimization framework based on accuracy-aware program transformations. These transformations reduce accuracy of computation in return for improved performance, energy efficiency, and/or resilience. The optimization framework includes program analyses that characterize the accuracy of transformed programs, and search techniques that navigate the tradeoff space induced by transformations to find approximate programs with profitable tradeoffs. Our research shows that this approach can significantly improve program performance while providing acceptably accurate results.

KW - Accuracy

KW - Approximation methods

KW - Computational efficiency

KW - Hardware

KW - Optimization

KW - Probabilistic logic

KW - Programming

UR - http://www.scopus.com/inward/record.url?scp=84962222578&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84962222578&partnerID=8YFLogxK

U2 - 10.1109/CASES.2015.7324543

DO - 10.1109/CASES.2015.7324543

M3 - Conference contribution

AN - SCOPUS:84962222578

T3 - 2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015

SP - 37

EP - 38

BT - 2015 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES 2015

PB - Institute of Electrical and Electronics Engineers Inc.

ER -