Implementation and performance analysis of data mining classification algorithms on smartphones

Zahid Islam, Darren Yates, Junbin Gao

Research output: Book chapter/Published conference paperConference paper

Abstract

Smartphones are increasingly being used to capture data and perform complex tasks, however, this rarely extends to the local training of data models.This study investigates the implementation of data mining classification algorithms on smartphones, using 20 classifiers and nine mixed-design datasets on three devices. Their accuracy and processing speed are further compared against a laptop computer using our cross-platform ‘DataBench’ software. Results show that smartphones not only deliver classification accuracy equal to that of more powerful computers when using the same algorithms, as expected, but also, as many as 75% of the 180 algorithm/dataset learning tasks tested were completed on smartphone hardware within three seconds. However, tests further show that the increased complexity of newer algorithms searching forever greater classification accuracy is resulting in model-build times growing at an exponential rate. Additional testing identified that while a single algorithm execution can have negligible effect on battery life, power efficiency is affected by algorithm complexity, data size and attribute type. The increased processing demand of local model-learning on smartphones also results in increased power dissipation. Yet, even on a continuous-loop execution basis, mobile temperaturegains over a 15-minute period did not exceed 7°C. Our conclusion is that smartphones are ready to form self-reliant mobile data-mining solutions able to efficiently execute a wide range of classification algorithms. This offers numerous advantages, including data security and privacy improvements, removal of reliance on network connectivity and delivery of personalised learning.
Original languageEnglish
Title of host publicationData Mining
Subtitle of host publication16th Australasian Conference, AusDM 2018 Bahrurst, NSW, Australia, November 28–30, 2018 Revised Selected Papers
EditorsRafiqul Islam, Yun Sing Koh, Yanchang Zhao, Graco Warwick, David Stirling, Chang-Tsun Li, Zuhidul Islam
Place of PublicationSingapore
PublisherSpringer Singapore
Pages331-343
Number of pages13
Volume996
ISBN (Electronic)9789811366611
ISBN (Print)9789811366604
Publication statusPublished - Feb 2019
Event16th Australasian Data Mining Conference (AusDM 2018) - Charles Sturt University, Bathurst, Australia
Duration: 28 Nov 201830 Nov 2018
https://www.springer.com/us/book/9789811366604 (link to conf proceedings book)
https://ausdm18.ausdm.org/ (conference website)

Conference

Conference16th Australasian Data Mining Conference (AusDM 2018)
CountryAustralia
CityBathurst
Period28/11/1830/11/18
OtherThe Australasian Data Mining Conference (AusDM) has established itself as the premier Australasian meeting for both practitioners and researchers in data mining. It is devoted to the art and science of intelligent analysis of (usually big) data sets for meaningful (and previously unknown) insights. This conference will enable the sharing and learning of research and progress in the local context and new breakthroughs in data mining algorithms and their applications across all industries.
Internet address

Fingerprint

Smartphones
Data mining
Laptop computers
Data privacy
Security of data
Processing
Learning algorithms
Data structures
Data acquisition
Energy dissipation
Classifiers
Hardware
Testing

Cite this

Islam, Z., Yates, D., & Gao, J. (2019). Implementation and performance analysis of data mining classification algorithms on smartphones. In R. Islam, Y. S. Koh, Y. Zhao, G. Warwick, D. Stirling, C-T. Li, & Z. Islam (Eds.), Data Mining: 16th Australasian Conference, AusDM 2018 Bahrurst, NSW, Australia, November 28–30, 2018 Revised Selected Papers (Vol. 996, pp. 331-343). [40] Singapore: Springer Singapore.
Islam, Zahid ; Yates, Darren ; Gao, Junbin. / Implementation and performance analysis of data mining classification algorithms on smartphones. Data Mining: 16th Australasian Conference, AusDM 2018 Bahrurst, NSW, Australia, November 28–30, 2018 Revised Selected Papers . editor / Rafiqul Islam ; Yun Sing Koh ; Yanchang Zhao ; Graco Warwick ; David Stirling ; Chang-Tsun Li ; Zuhidul Islam. Vol. 996 Singapore : Springer Singapore, 2019. pp. 331-343
@inproceedings{f4ed51479086491dbcb62e6f429ea90b,
title = "Implementation and performance analysis of data mining classification algorithms on smartphones",
abstract = "Smartphones are increasingly being used to capture data and perform complex tasks, however, this rarely extends to the local training of data models.This study investigates the implementation of data mining classification algorithms on smartphones, using 20 classifiers and nine mixed-design datasets on three devices. Their accuracy and processing speed are further compared against a laptop computer using our cross-platform ‘DataBench’ software. Results show that smartphones not only deliver classification accuracy equal to that of more powerful computers when using the same algorithms, as expected, but also, as many as 75{\%} of the 180 algorithm/dataset learning tasks tested were completed on smartphone hardware within three seconds. However, tests further show that the increased complexity of newer algorithms searching forever greater classification accuracy is resulting in model-build times growing at an exponential rate. Additional testing identified that while a single algorithm execution can have negligible effect on battery life, power efficiency is affected by algorithm complexity, data size and attribute type. The increased processing demand of local model-learning on smartphones also results in increased power dissipation. Yet, even on a continuous-loop execution basis, mobile temperaturegains over a 15-minute period did not exceed 7°C. Our conclusion is that smartphones are ready to form self-reliant mobile data-mining solutions able to efficiently execute a wide range of classification algorithms. This offers numerous advantages, including data security and privacy improvements, removal of reliance on network connectivity and delivery of personalised learning.",
author = "Zahid Islam and Darren Yates and Junbin Gao",
year = "2019",
month = "2",
language = "English",
isbn = "9789811366604",
volume = "996",
pages = "331--343",
editor = "Rafiqul Islam and Koh, {Yun Sing} and Yanchang Zhao and Warwick, {Graco } and Stirling, {David } and Chang-Tsun Li and Zuhidul Islam",
booktitle = "Data Mining",
publisher = "Springer Singapore",

}

Islam, Z, Yates, D & Gao, J 2019, Implementation and performance analysis of data mining classification algorithms on smartphones. in R Islam, YS Koh, Y Zhao, G Warwick, D Stirling, C-T Li & Z Islam (eds), Data Mining: 16th Australasian Conference, AusDM 2018 Bahrurst, NSW, Australia, November 28–30, 2018 Revised Selected Papers . vol. 996, 40, Springer Singapore, Singapore, pp. 331-343, 16th Australasian Data Mining Conference (AusDM 2018), Bathurst, Australia, 28/11/18.

Implementation and performance analysis of data mining classification algorithms on smartphones. / Islam, Zahid; Yates, Darren; Gao, Junbin.

Data Mining: 16th Australasian Conference, AusDM 2018 Bahrurst, NSW, Australia, November 28–30, 2018 Revised Selected Papers . ed. / Rafiqul Islam; Yun Sing Koh; Yanchang Zhao; Graco Warwick; David Stirling; Chang-Tsun Li; Zuhidul Islam. Vol. 996 Singapore : Springer Singapore, 2019. p. 331-343 40.

Research output: Book chapter/Published conference paperConference paper

TY - GEN

T1 - Implementation and performance analysis of data mining classification algorithms on smartphones

AU - Islam, Zahid

AU - Yates, Darren

AU - Gao, Junbin

PY - 2019/2

Y1 - 2019/2

N2 - Smartphones are increasingly being used to capture data and perform complex tasks, however, this rarely extends to the local training of data models.This study investigates the implementation of data mining classification algorithms on smartphones, using 20 classifiers and nine mixed-design datasets on three devices. Their accuracy and processing speed are further compared against a laptop computer using our cross-platform ‘DataBench’ software. Results show that smartphones not only deliver classification accuracy equal to that of more powerful computers when using the same algorithms, as expected, but also, as many as 75% of the 180 algorithm/dataset learning tasks tested were completed on smartphone hardware within three seconds. However, tests further show that the increased complexity of newer algorithms searching forever greater classification accuracy is resulting in model-build times growing at an exponential rate. Additional testing identified that while a single algorithm execution can have negligible effect on battery life, power efficiency is affected by algorithm complexity, data size and attribute type. The increased processing demand of local model-learning on smartphones also results in increased power dissipation. Yet, even on a continuous-loop execution basis, mobile temperaturegains over a 15-minute period did not exceed 7°C. Our conclusion is that smartphones are ready to form self-reliant mobile data-mining solutions able to efficiently execute a wide range of classification algorithms. This offers numerous advantages, including data security and privacy improvements, removal of reliance on network connectivity and delivery of personalised learning.

AB - Smartphones are increasingly being used to capture data and perform complex tasks, however, this rarely extends to the local training of data models.This study investigates the implementation of data mining classification algorithms on smartphones, using 20 classifiers and nine mixed-design datasets on three devices. Their accuracy and processing speed are further compared against a laptop computer using our cross-platform ‘DataBench’ software. Results show that smartphones not only deliver classification accuracy equal to that of more powerful computers when using the same algorithms, as expected, but also, as many as 75% of the 180 algorithm/dataset learning tasks tested were completed on smartphone hardware within three seconds. However, tests further show that the increased complexity of newer algorithms searching forever greater classification accuracy is resulting in model-build times growing at an exponential rate. Additional testing identified that while a single algorithm execution can have negligible effect on battery life, power efficiency is affected by algorithm complexity, data size and attribute type. The increased processing demand of local model-learning on smartphones also results in increased power dissipation. Yet, even on a continuous-loop execution basis, mobile temperaturegains over a 15-minute period did not exceed 7°C. Our conclusion is that smartphones are ready to form self-reliant mobile data-mining solutions able to efficiently execute a wide range of classification algorithms. This offers numerous advantages, including data security and privacy improvements, removal of reliance on network connectivity and delivery of personalised learning.

M3 - Conference paper

SN - 9789811366604

VL - 996

SP - 331

EP - 343

BT - Data Mining

A2 - Islam, Rafiqul

A2 - Koh, Yun Sing

A2 - Zhao, Yanchang

A2 - Warwick, Graco

A2 - Stirling, David

A2 - Li, Chang-Tsun

A2 - Islam, Zuhidul

PB - Springer Singapore

CY - Singapore

ER -

Islam Z, Yates D, Gao J. Implementation and performance analysis of data mining classification algorithms on smartphones. In Islam R, Koh YS, Zhao Y, Warwick G, Stirling D, Li C-T, Islam Z, editors, Data Mining: 16th Australasian Conference, AusDM 2018 Bahrurst, NSW, Australia, November 28–30, 2018 Revised Selected Papers . Vol. 996. Singapore: Springer Singapore. 2019. p. 331-343. 40