On the Perceptron’s Compression

Shay Moran; Ido Nachum; Itai Panasoff; Amir Yehudayoff

doi:10.1007/978-3-030-51466-2_29

On the Perceptron’s Compression

Shay Moran, Ido Nachum, Itai Panasoff, Amir Yehudayoff

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

Abstract

We study and provide exposition to several phenomena that are related to the perceptron’s compression. One theme concerns modifications of the perceptron algorithm that yield better guarantees on the margin of the hyperplane it outputs. These modifications can be useful in training neural networks as well, and we demonstrate them with some experimental data. In a second theme, we deduce conclusions from the perceptron’s compression in various contexts.

Original language	English
Title of host publication	Beyond the Horizon of Computability - 16th Conference on Computability in Europe, CiE 2020, Proceedings
Editors	Marcella Anselmo, Gianluca Della Vedova, Florin Manea, Arno Pauly
Publisher	Springer
Pages	310-325
Number of pages	16
ISBN (Print)	9783030514655
DOIs	https://doi.org/10.1007/978-3-030-51466-2_29
State	Published - 2020
Externally published	Yes
Event	16th Conference on Computability in Europe, CiE 2020 - Fisciano, Italy Duration: 29 Jun 2020 → 3 Jul 2020

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12098 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th Conference on Computability in Europe, CiE 2020
Country/Territory	Italy
City	Fisciano
Period	29/06/20 → 3/07/20

Bibliographical note

Publisher Copyright:
© 2020, Springer Nature Switzerland AG.

Keywords

Compression
Convex separation
Machine learning

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-030-51466-2_29

Cite this

Moran, S., Nachum, I., Panasoff, I., & Yehudayoff, A. (2020). On the Perceptron’s Compression. In M. Anselmo, G. Della Vedova, F. Manea, & A. Pauly (Eds.), Beyond the Horizon of Computability - 16th Conference on Computability in Europe, CiE 2020, Proceedings (pp. 310-325). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12098 LNCS). Springer. https://doi.org/10.1007/978-3-030-51466-2_29

Moran, Shay ; Nachum, Ido ; Panasoff, Itai et al. / On the Perceptron’s Compression. Beyond the Horizon of Computability - 16th Conference on Computability in Europe, CiE 2020, Proceedings. editor / Marcella Anselmo ; Gianluca Della Vedova ; Florin Manea ; Arno Pauly. Springer, 2020. pp. 310-325 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{5f0b9ddb84934a1289b26cd2e4e527fb,

title = "On the Perceptron{\textquoteright}s Compression",

abstract = "We study and provide exposition to several phenomena that are related to the perceptron{\textquoteright}s compression. One theme concerns modifications of the perceptron algorithm that yield better guarantees on the margin of the hyperplane it outputs. These modifications can be useful in training neural networks as well, and we demonstrate them with some experimental data. In a second theme, we deduce conclusions from the perceptron{\textquoteright}s compression in various contexts.",

keywords = "Compression, Convex separation, Machine learning",

author = "Shay Moran and Ido Nachum and Itai Panasoff and Amir Yehudayoff",

note = "Publisher Copyright: {\textcopyright} 2020, Springer Nature Switzerland AG.; 16th Conference on Computability in Europe, CiE 2020 ; Conference date: 29-06-2020 Through 03-07-2020",

year = "2020",

doi = "10.1007/978-3-030-51466-2_29",

language = "English",

isbn = "9783030514655",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "310--325",

editor = "Marcella Anselmo and {Della Vedova}, Gianluca and Florin Manea and Arno Pauly",

booktitle = "Beyond the Horizon of Computability - 16th Conference on Computability in Europe, CiE 2020, Proceedings",

address = "Netherlands",

}

Moran, S, Nachum, I, Panasoff, I & Yehudayoff, A 2020, On the Perceptron’s Compression. in M Anselmo, G Della Vedova, F Manea & A Pauly (eds), Beyond the Horizon of Computability - 16th Conference on Computability in Europe, CiE 2020, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12098 LNCS, Springer, pp. 310-325, 16th Conference on Computability in Europe, CiE 2020, Fisciano, Italy, 29/06/20. https://doi.org/10.1007/978-3-030-51466-2_29

On the Perceptron’s Compression. / Moran, Shay; Nachum, Ido; Panasoff, Itai et al.
Beyond the Horizon of Computability - 16th Conference on Computability in Europe, CiE 2020, Proceedings. ed. / Marcella Anselmo; Gianluca Della Vedova; Florin Manea; Arno Pauly. Springer, 2020. p. 310-325 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12098 LNCS).

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

TY - GEN

T1 - On the Perceptron’s Compression

AU - Moran, Shay

AU - Nachum, Ido

AU - Panasoff, Itai

AU - Yehudayoff, Amir

PY - 2020

Y1 - 2020

N2 - We study and provide exposition to several phenomena that are related to the perceptron’s compression. One theme concerns modifications of the perceptron algorithm that yield better guarantees on the margin of the hyperplane it outputs. These modifications can be useful in training neural networks as well, and we demonstrate them with some experimental data. In a second theme, we deduce conclusions from the perceptron’s compression in various contexts.

AB - We study and provide exposition to several phenomena that are related to the perceptron’s compression. One theme concerns modifications of the perceptron algorithm that yield better guarantees on the margin of the hyperplane it outputs. These modifications can be useful in training neural networks as well, and we demonstrate them with some experimental data. In a second theme, we deduce conclusions from the perceptron’s compression in various contexts.

KW - Compression

KW - Convex separation

KW - Machine learning

UR - http://www.scopus.com/inward/record.url?scp=85088211475&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-51466-2_29

DO - 10.1007/978-3-030-51466-2_29

M3 - Conference contribution

AN - SCOPUS:85088211475

SN - 9783030514655

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 310

EP - 325

BT - Beyond the Horizon of Computability - 16th Conference on Computability in Europe, CiE 2020, Proceedings

A2 - Anselmo, Marcella

A2 - Della Vedova, Gianluca

A2 - Manea, Florin

A2 - Pauly, Arno

PB - Springer

T2 - 16th Conference on Computability in Europe, CiE 2020

Y2 - 29 June 2020 through 3 July 2020

ER -

Moran S, Nachum I, Panasoff I, Yehudayoff A. On the Perceptron’s Compression. In Anselmo M, Della Vedova G, Manea F, Pauly A, editors, Beyond the Horizon of Computability - 16th Conference on Computability in Europe, CiE 2020, Proceedings. Springer. 2020. p. 310-325. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-51466-2_29

On the Perceptron’s Compression

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this