Optimal Synthesis of IDK-Cascades

Sanjoy Baruah; Alan Burns; Yue Wu

doi:10.1145/3453417.3453425

Optimal Synthesis of IDK-Cascades

Sanjoy Baruah, Alan Burns, Yue Wu

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

A classifier is a software component, often based upon deep learning (DL), that categorizes each input provided to it into one of a fixed set of classes. An IDK classifier may additionally output an 'I don't know' (IDK) on certain input. Given several different IDK classifiers for the same operation, the problem is considered of using them in concert in such a manner that the average duration to successfully classify any input is minimized. Optimal algorithms are proposed for solving this problem, both as is and under an additional constraint that the operation must be completed within a specified hard deadline.

Original language	English
Title of host publication	RTNS'2021: 29th International Conference on Real-Time Networks and Systems
Publisher	ACM
Pages	184-191
Number of pages	8
ISBN (Print)	9781450390019
DOIs	https://doi.org/10.1145/3453417.3453425
Publication status	Published - 22 Jul 2021
Event	Real-Time Networked Systems - Duration: 7 Apr 2021 → 9 Apr 2021

Conference

Conference	Real-Time Networked Systems
Abbreviated title	RTNS
Period	7/04/21 → 9/04/21

Bibliographical note

© 2021 Association for Computing Machinery. This is an author-produced version of the published paper. Uploaded in accordance with the publisher’s self-archiving policy. Further copying may not be permitted; contact the publisher for details

Access to Document

10.1145/3453417.3453425Licence: Unspecified

00main
© 2021 Association for Computing Machinery. This is an author-produced version of the published paper. Uploaded in accordance with the publisher’s self-archiving policy. Further copying may not be permitted; contact the publisher for details
Accepted author manuscript, 767 KBLicence: Unspecified

Cite this

@inproceedings{e048a68aaa3a4d6ab771cb33c4b55073,

title = "Optimal Synthesis of IDK-Cascades",

abstract = "A classifier is a software component, often based upon deep learning (DL), that categorizes each input provided to it into one of a fixed set of classes. An IDK classifier may additionally output an 'I don't know' (IDK) on certain input. Given several different IDK classifiers for the same operation, the problem is considered of using them in concert in such a manner that the average duration to successfully classify any input is minimized. Optimal algorithms are proposed for solving this problem, both as is and under an additional constraint that the operation must be completed within a specified hard deadline.",

author = "Sanjoy Baruah and Alan Burns and Yue Wu",

note = "{\textcopyright} 2021 Association for Computing Machinery. This is an author-produced version of the published paper. Uploaded in accordance with the publisher{\textquoteright}s self-archiving policy. Further copying may not be permitted; contact the publisher for details ; Real-Time Networked Systems, RTNS ; Conference date: 07-04-2021 Through 09-04-2021",

year = "2021",

month = jul,

day = "22",

doi = "10.1145/3453417.3453425",

language = "English",

isbn = "9781450390019",

pages = "184--191",

booktitle = "RTNS'2021: 29th International Conference on Real-Time Networks and Systems",

publisher = "ACM",

address = "United States",

}

TY - GEN

T1 - Optimal Synthesis of IDK-Cascades

AU - Baruah, Sanjoy

AU - Burns, Alan

AU - Wu, Yue

N1 - © 2021 Association for Computing Machinery. This is an author-produced version of the published paper. Uploaded in accordance with the publisher’s self-archiving policy. Further copying may not be permitted; contact the publisher for details

PY - 2021/7/22

Y1 - 2021/7/22

N2 - A classifier is a software component, often based upon deep learning (DL), that categorizes each input provided to it into one of a fixed set of classes. An IDK classifier may additionally output an 'I don't know' (IDK) on certain input. Given several different IDK classifiers for the same operation, the problem is considered of using them in concert in such a manner that the average duration to successfully classify any input is minimized. Optimal algorithms are proposed for solving this problem, both as is and under an additional constraint that the operation must be completed within a specified hard deadline.

AB - A classifier is a software component, often based upon deep learning (DL), that categorizes each input provided to it into one of a fixed set of classes. An IDK classifier may additionally output an 'I don't know' (IDK) on certain input. Given several different IDK classifiers for the same operation, the problem is considered of using them in concert in such a manner that the average duration to successfully classify any input is minimized. Optimal algorithms are proposed for solving this problem, both as is and under an additional constraint that the operation must be completed within a specified hard deadline.

U2 - 10.1145/3453417.3453425

DO - 10.1145/3453417.3453425

M3 - Conference contribution

SN - 9781450390019

SP - 184

EP - 191

BT - RTNS'2021: 29th International Conference on Real-Time Networks and Systems

PB - ACM

T2 - Real-Time Networked Systems

Y2 - 7 April 2021 through 9 April 2021

ER -