Fast and robust supervised learning in high dimensions using the geometry of the data

Ujjal Kumar Mukherjee, Subhabrata Majumdar, Snigdhansu Chatterjee

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

Abstract

We develop a method for tracing out the shape of a cloud of sample observations, in arbitrary dimensions, called the data cloud wrapper (DCW). The DCW have strong theoretical properties, have algorithmic scalability and parallel computational features. We further use the DCW to develop a new fast, robust and accurate classification method in high dimensions, called the geometric learning algorithm (GLA). Two of the main features of the proposed algorithm are that there are no assumptions made about the geometric properties of the underlying data generating distribution, and that there are no parametric or other restrictive assumptions made either for the data or the algorithm. The proposed methods are typically faster and more robust than established classification techniques, while being comparably accurate in most cases.

Original languageEnglish (US)
Title of host publicationAdvances in Data Mining
Subtitle of host publicationApplications and Theoretical Aspects - 15th Industrial Conference, ICDM 2015, Proceedings
EditorsPetra Perner
PublisherSpringer- Verlag
Pages109-123
Number of pages15
ISBN (Print)9783319209098
DOIs
StatePublished - Jan 1 2015
Event15th Industrial Conference on Data Mining, ICDM 2015 - Hamburg, Germany
Duration: Jul 11 2015Jul 24 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9165
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other15th Industrial Conference on Data Mining, ICDM 2015
CountryGermany
CityHamburg
Period7/11/157/24/15

Fingerprint

Supervised learning
Supervised Learning
Higher Dimensions
Wrapper
Geometry
Learning algorithms
Scalability
Geometric Algorithms
Tracing
Learning Algorithm
Arbitrary

Cite this

Mukherjee, U. K., Majumdar, S., & Chatterjee, S. (2015). Fast and robust supervised learning in high dimensions using the geometry of the data. In P. Perner (Ed.), Advances in Data Mining: Applications and Theoretical Aspects - 15th Industrial Conference, ICDM 2015, Proceedings (pp. 109-123). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9165). Springer- Verlag. https://doi.org/10.1007/978-3-319-20910-4_9

Fast and robust supervised learning in high dimensions using the geometry of the data. / Mukherjee, Ujjal Kumar; Majumdar, Subhabrata; Chatterjee, Snigdhansu.

Advances in Data Mining: Applications and Theoretical Aspects - 15th Industrial Conference, ICDM 2015, Proceedings. ed. / Petra Perner. Springer- Verlag, 2015. p. 109-123 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9165).

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

Mukherjee, UK, Majumdar, S & Chatterjee, S 2015, Fast and robust supervised learning in high dimensions using the geometry of the data. in P Perner (ed.), Advances in Data Mining: Applications and Theoretical Aspects - 15th Industrial Conference, ICDM 2015, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9165, Springer- Verlag, pp. 109-123, 15th Industrial Conference on Data Mining, ICDM 2015, Hamburg, Germany, 7/11/15. https://doi.org/10.1007/978-3-319-20910-4_9
Mukherjee UK, Majumdar S, Chatterjee S. Fast and robust supervised learning in high dimensions using the geometry of the data. In Perner P, editor, Advances in Data Mining: Applications and Theoretical Aspects - 15th Industrial Conference, ICDM 2015, Proceedings. Springer- Verlag. 2015. p. 109-123. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-20910-4_9
Mukherjee, Ujjal Kumar ; Majumdar, Subhabrata ; Chatterjee, Snigdhansu. / Fast and robust supervised learning in high dimensions using the geometry of the data. Advances in Data Mining: Applications and Theoretical Aspects - 15th Industrial Conference, ICDM 2015, Proceedings. editor / Petra Perner. Springer- Verlag, 2015. pp. 109-123 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{823f21ce95004188b464e179a0036d3a,
title = "Fast and robust supervised learning in high dimensions using the geometry of the data",
abstract = "We develop a method for tracing out the shape of a cloud of sample observations, in arbitrary dimensions, called the data cloud wrapper (DCW). The DCW have strong theoretical properties, have algorithmic scalability and parallel computational features. We further use the DCW to develop a new fast, robust and accurate classification method in high dimensions, called the geometric learning algorithm (GLA). Two of the main features of the proposed algorithm are that there are no assumptions made about the geometric properties of the underlying data generating distribution, and that there are no parametric or other restrictive assumptions made either for the data or the algorithm. The proposed methods are typically faster and more robust than established classification techniques, while being comparably accurate in most cases.",
author = "Mukherjee, {Ujjal Kumar} and Subhabrata Majumdar and Snigdhansu Chatterjee",
year = "2015",
month = "1",
day = "1",
doi = "10.1007/978-3-319-20910-4_9",
language = "English (US)",
isbn = "9783319209098",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer- Verlag",
pages = "109--123",
editor = "Petra Perner",
booktitle = "Advances in Data Mining",

}

TY - GEN

T1 - Fast and robust supervised learning in high dimensions using the geometry of the data

AU - Mukherjee, Ujjal Kumar

AU - Majumdar, Subhabrata

AU - Chatterjee, Snigdhansu

PY - 2015/1/1

Y1 - 2015/1/1

N2 - We develop a method for tracing out the shape of a cloud of sample observations, in arbitrary dimensions, called the data cloud wrapper (DCW). The DCW have strong theoretical properties, have algorithmic scalability and parallel computational features. We further use the DCW to develop a new fast, robust and accurate classification method in high dimensions, called the geometric learning algorithm (GLA). Two of the main features of the proposed algorithm are that there are no assumptions made about the geometric properties of the underlying data generating distribution, and that there are no parametric or other restrictive assumptions made either for the data or the algorithm. The proposed methods are typically faster and more robust than established classification techniques, while being comparably accurate in most cases.

AB - We develop a method for tracing out the shape of a cloud of sample observations, in arbitrary dimensions, called the data cloud wrapper (DCW). The DCW have strong theoretical properties, have algorithmic scalability and parallel computational features. We further use the DCW to develop a new fast, robust and accurate classification method in high dimensions, called the geometric learning algorithm (GLA). Two of the main features of the proposed algorithm are that there are no assumptions made about the geometric properties of the underlying data generating distribution, and that there are no parametric or other restrictive assumptions made either for the data or the algorithm. The proposed methods are typically faster and more robust than established classification techniques, while being comparably accurate in most cases.

UR - http://www.scopus.com/inward/record.url?scp=84950121438&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84950121438&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-20910-4_9

DO - 10.1007/978-3-319-20910-4_9

M3 - Conference contribution

AN - SCOPUS:84950121438

SN - 9783319209098

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

SP - 109

EP - 123

BT - Advances in Data Mining

A2 - Perner, Petra

PB - Springer- Verlag

ER -