CIM: A reliable metric for evaluating program phase classifications

Sreekumar V. Kodakara; Jinpyo Kim; David J. Lilja; Douglas Hawkins; Wei Chung Hsut; Pen Chung Yew

doi:10.1109/L-CA.2007.4

CIM: A reliable metric for evaluating program phase classifications

Sreekumar V. Kodakara, Jinpyo Kim, David J. Lilja, Douglas Hawkins, Wei Chung Hsut, Pen Chung Yew

Research output: Contribution to journal › Article › peer-review

5 Scopus citations

Abstract

We propose the use of the Confidence Interval of estimated Mean (CIM), a metric based on statistical sampling theory, to evaluate the quality of a given phase classification and for comparing different phase classification schemes. Previous research on phase classification used the Weighted Average of Coefficient of Variation (CoVwa) to estimate phase classification quality. We found that the phase quality indicated by CoV_wa could be inconsistent across different phase classifications. We explain the reasons behind this inconsistency and demonstrate the inconsistency using data from several SPEC CPU2000 benchmark programs. We show that the Confidence Interval of estimated Mean (CIM) correctly estimates the quality of phase classification with a meaningful statistical interpretation.

Original language	English (US)
Pages (from-to)	9-12
Number of pages	4
Journal	IEEE Computer Architecture Letters
Volume	6
Issue number	1
DOIs	https://doi.org/10.1109/L-CA.2007.4
State	Published - Jan 2007

Bibliographical note

Funding Information:
VI. ACKNOWLEDGEMENT This work was supported in part by grant NSF no. CCF-0541162, EIA-0220021, Intel, the University of Minnesota Digital Technology Center, and the Minnesota Supercomputing Institute. REFERENCES

Access

10.1109/L-CA.2007.4

OpenUrl availability

Full text

Cite this

@article{9de4422adeb94f2b81b7caa7c6b2957d,

title = "CIM: A reliable metric for evaluating program phase classifications",

abstract = "We propose the use of the Confidence Interval of estimated Mean (CIM), a metric based on statistical sampling theory, to evaluate the quality of a given phase classification and for comparing different phase classification schemes. Previous research on phase classification used the Weighted Average of Coefficient of Variation (CoVwa) to estimate phase classification quality. We found that the phase quality indicated by CoVwa could be inconsistent across different phase classifications. We explain the reasons behind this inconsistency and demonstrate the inconsistency using data from several SPEC CPU2000 benchmark programs. We show that the Confidence Interval of estimated Mean (CIM) correctly estimates the quality of phase classification with a meaningful statistical interpretation.",

author = "Kodakara, {Sreekumar V.} and Jinpyo Kim and Lilja, {David J.} and Douglas Hawkins and Hsut, {Wei Chung} and Yew, {Pen Chung}",

note = "Funding Information: VI. ACKNOWLEDGEMENT This work was supported in part by grant NSF no. CCF-0541162, EIA-0220021, Intel, the University of Minnesota Digital Technology Center, and the Minnesota Supercomputing Institute. REFERENCES",

year = "2007",

month = jan,

doi = "10.1109/L-CA.2007.4",

language = "English (US)",

volume = "6",

pages = "9--12",

journal = "IEEE Computer Architecture Letters",

issn = "1556-6056",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "1",

}

TY - JOUR

T1 - CIM

T2 - A reliable metric for evaluating program phase classifications

AU - Kodakara, Sreekumar V.

AU - Kim, Jinpyo

AU - Lilja, David J.

AU - Hawkins, Douglas

AU - Hsut, Wei Chung

AU - Yew, Pen Chung

N1 - Funding Information: VI. ACKNOWLEDGEMENT This work was supported in part by grant NSF no. CCF-0541162, EIA-0220021, Intel, the University of Minnesota Digital Technology Center, and the Minnesota Supercomputing Institute. REFERENCES

PY - 2007/1

Y1 - 2007/1

N2 - We propose the use of the Confidence Interval of estimated Mean (CIM), a metric based on statistical sampling theory, to evaluate the quality of a given phase classification and for comparing different phase classification schemes. Previous research on phase classification used the Weighted Average of Coefficient of Variation (CoVwa) to estimate phase classification quality. We found that the phase quality indicated by CoVwa could be inconsistent across different phase classifications. We explain the reasons behind this inconsistency and demonstrate the inconsistency using data from several SPEC CPU2000 benchmark programs. We show that the Confidence Interval of estimated Mean (CIM) correctly estimates the quality of phase classification with a meaningful statistical interpretation.

AB - We propose the use of the Confidence Interval of estimated Mean (CIM), a metric based on statistical sampling theory, to evaluate the quality of a given phase classification and for comparing different phase classification schemes. Previous research on phase classification used the Weighted Average of Coefficient of Variation (CoVwa) to estimate phase classification quality. We found that the phase quality indicated by CoVwa could be inconsistent across different phase classifications. We explain the reasons behind this inconsistency and demonstrate the inconsistency using data from several SPEC CPU2000 benchmark programs. We show that the Confidence Interval of estimated Mean (CIM) correctly estimates the quality of phase classification with a meaningful statistical interpretation.

UR - http://www.scopus.com/inward/record.url?scp=34547641909&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34547641909&partnerID=8YFLogxK

U2 - 10.1109/L-CA.2007.4

DO - 10.1109/L-CA.2007.4

M3 - Article

AN - SCOPUS:34547641909

SN - 1556-6056

VL - 6

SP - 9

EP - 12

JO - IEEE Computer Architecture Letters

JF - IEEE Computer Architecture Letters

IS - 1

ER -

CIM: A reliable metric for evaluating program phase classifications

Abstract

Bibliographical note

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this