Gravity inspired clustering algorithm

Imran Mohammed; Iain B. Collings; Stephen V. Hanly

doi:10.1109/ICSPCS50536.2020.9310038

Gravity inspired clustering algorithm

Imran Mohammed, Iain B. Collings, Stephen V. Hanly

School of Engineering

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

1 Citation (Scopus)

Abstract

This paper presents a new clustering algorithm inspired by Newtonian gravity that iteratively groups data and eliminates outliers. In particular, we impose a grid over the region of interest and define a particle with data-dependent mass for each grid square. We then calculate a Newtonian inspired force on each of the particles and move them in the direction of the force. We repeat the process until there is no further movement. We compare performance with existing algorithms and show that in cases of medium to high clutter, our algorithm has an order of magnitude lower estimation error.

Original language	English
Title of host publication	2020, 14th International Conference on Signal Processing and Communication Systems, (ICSPCS)
Subtitle of host publication	proceedings
Editors	Tadeusz A. Wysocki, Beata J. Wysocki
Place of Publication	Piscataway, NJ
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Number of pages	6
ISBN (Electronic)	9781728199726, 9781728199719
ISBN (Print)	9781728199733
DOIs	https://doi.org/10.1109/ICSPCS50536.2020.9310038
Publication status	Published - 2020
Event	14th International Conference on Signal Processing and Communication Systems, ICSPCS 2020 - Virtual Duration: 14 Dec 2020 → 16 Dec 2020

Conference

Conference	14th International Conference on Signal Processing and Communication Systems, ICSPCS 2020
City	Virtual
Period	14/12/20 → 16/12/20

Access to Document

10.1109/ICSPCS50536.2020.9310038

Cite this

@inproceedings{8a399e5c053a4ff0916846463c16dd3b,

title = "Gravity inspired clustering algorithm",

abstract = "This paper presents a new clustering algorithm inspired by Newtonian gravity that iteratively groups data and eliminates outliers. In particular, we impose a grid over the region of interest and define a particle with data-dependent mass for each grid square. We then calculate a Newtonian inspired force on each of the particles and move them in the direction of the force. We repeat the process until there is no further movement. We compare performance with existing algorithms and show that in cases of medium to high clutter, our algorithm has an order of magnitude lower estimation error.",

author = "Imran Mohammed and Collings, {Iain B.} and Hanly, {Stephen V.}",

year = "2020",

doi = "10.1109/ICSPCS50536.2020.9310038",

language = "English",

isbn = "9781728199733",

editor = "Wysocki, {Tadeusz A.} and Wysocki, {Beata J.}",

booktitle = "2020, 14th International Conference on Signal Processing and Communication Systems, (ICSPCS)",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

address = "United States",

note = "14th International Conference on Signal Processing and Communication Systems, ICSPCS 2020 ; Conference date: 14-12-2020 Through 16-12-2020",

}

Mohammed, I, Collings, IB & Hanly, SV 2020, Gravity inspired clustering algorithm. in TA Wysocki & BJ Wysocki (eds), 2020, 14th International Conference on Signal Processing and Communication Systems, (ICSPCS): proceedings. Institute of Electrical and Electronics Engineers (IEEE), Piscataway, NJ, 14th International Conference on Signal Processing and Communication Systems, ICSPCS 2020, Virtual, 14/12/20. https://doi.org/10.1109/ICSPCS50536.2020.9310038

Gravity inspired clustering algorithm. / Mohammed, Imran; Collings, Iain B.; Hanly, Stephen V.
2020, 14th International Conference on Signal Processing and Communication Systems, (ICSPCS): proceedings. ed. / Tadeusz A. Wysocki; Beata J. Wysocki. Piscataway, NJ: Institute of Electrical and Electronics Engineers (IEEE), 2020.

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

TY - GEN

T1 - Gravity inspired clustering algorithm

AU - Mohammed, Imran

AU - Collings, Iain B.

AU - Hanly, Stephen V.

PY - 2020

Y1 - 2020

N2 - This paper presents a new clustering algorithm inspired by Newtonian gravity that iteratively groups data and eliminates outliers. In particular, we impose a grid over the region of interest and define a particle with data-dependent mass for each grid square. We then calculate a Newtonian inspired force on each of the particles and move them in the direction of the force. We repeat the process until there is no further movement. We compare performance with existing algorithms and show that in cases of medium to high clutter, our algorithm has an order of magnitude lower estimation error.

AB - This paper presents a new clustering algorithm inspired by Newtonian gravity that iteratively groups data and eliminates outliers. In particular, we impose a grid over the region of interest and define a particle with data-dependent mass for each grid square. We then calculate a Newtonian inspired force on each of the particles and move them in the direction of the force. We repeat the process until there is no further movement. We compare performance with existing algorithms and show that in cases of medium to high clutter, our algorithm has an order of magnitude lower estimation error.

UR - http://www.scopus.com/inward/record.url?scp=85100033426&partnerID=8YFLogxK

U2 - 10.1109/ICSPCS50536.2020.9310038

DO - 10.1109/ICSPCS50536.2020.9310038

M3 - Conference proceeding contribution

AN - SCOPUS:85100033426

SN - 9781728199733

BT - 2020, 14th International Conference on Signal Processing and Communication Systems, (ICSPCS)

A2 - Wysocki, Tadeusz A.

A2 - Wysocki, Beata J.

PB - Institute of Electrical and Electronics Engineers (IEEE)

CY - Piscataway, NJ

T2 - 14th International Conference on Signal Processing and Communication Systems, ICSPCS 2020

Y2 - 14 December 2020 through 16 December 2020

ER -

Gravity inspired clustering algorithm

Abstract

Conference

Access to Document

Other files and links

Fingerprint

Cite this