Fast knn classification based on softcore cpu and reconfigurable hardware

Hui Ya Li, Yao Jung Yeh, Wen Jyi Hwang

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

This paper presents a novel architecture for k-nearest neighbor (kNN) classification using field programmable gate array (FPGA). In the architecture, the first k closest vectors in the design set of a kNN classifier for each input vector are first identified by perfomung the partial distance search (PDS) in the wavelet domain. To implement the PDS in hardware, subspace search, bitplane reduction, multiple-coefficient accumulation and multiple-module computation techniques are employed for the effective reduction of the area complexity and computation latency. The proposed implementation has been embedded in a softcore CPU for physical performance measurement. Experimental results show that the implementation provides acosteffective solution to the FPGA realization of kNN classification systems where both high throughput and low area cost are desired.

Original languageEnglish
Pages (from-to)431-446
Number of pages16
JournalIntelligent Automation and Soft Computing
Volume17
Issue number4
DOIs
Publication statusPublished - 2011 Jan

Fingerprint

Reconfigurable hardware
Reconfigurable Hardware
Field programmable gate arrays (FPGA)
Nearest Neighbor
Field Programmable Gate Array
Program processors
Partial
Classifiers
Performance Measurement
Throughput
Hardware
High Throughput
Latency
Wavelets
Classifier
Subspace
Costs
Module
Experimental Results
Coefficient

Keywords

  • FPGA
  • Knn
  • Partial distance search
  • Reconfigurable computing

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Computational Theory and Mathematics
  • Artificial Intelligence

Cite this

Fast knn classification based on softcore cpu and reconfigurable hardware. / Li, Hui Ya; Yeh, Yao Jung; Hwang, Wen Jyi.

In: Intelligent Automation and Soft Computing, Vol. 17, No. 4, 01.2011, p. 431-446.

Research output: Contribution to journalArticle

@article{45e24ad19c51405fae01d04c3170e521,
title = "Fast knn classification based on softcore cpu and reconfigurable hardware",
abstract = "This paper presents a novel architecture for k-nearest neighbor (kNN) classification using field programmable gate array (FPGA). In the architecture, the first k closest vectors in the design set of a kNN classifier for each input vector are first identified by perfomung the partial distance search (PDS) in the wavelet domain. To implement the PDS in hardware, subspace search, bitplane reduction, multiple-coefficient accumulation and multiple-module computation techniques are employed for the effective reduction of the area complexity and computation latency. The proposed implementation has been embedded in a softcore CPU for physical performance measurement. Experimental results show that the implementation provides acosteffective solution to the FPGA realization of kNN classification systems where both high throughput and low area cost are desired.",
keywords = "FPGA, Knn, Partial distance search, Reconfigurable computing",
author = "Li, {Hui Ya} and Yeh, {Yao Jung} and Hwang, {Wen Jyi}",
year = "2011",
month = "1",
doi = "10.1080/10798587.2011.10643159",
language = "English",
volume = "17",
pages = "431--446",
journal = "Intelligent Automation and Soft Computing",
issn = "1079-8587",
publisher = "AutoSoft Press",
number = "4",

}

TY - JOUR

T1 - Fast knn classification based on softcore cpu and reconfigurable hardware

AU - Li, Hui Ya

AU - Yeh, Yao Jung

AU - Hwang, Wen Jyi

PY - 2011/1

Y1 - 2011/1

N2 - This paper presents a novel architecture for k-nearest neighbor (kNN) classification using field programmable gate array (FPGA). In the architecture, the first k closest vectors in the design set of a kNN classifier for each input vector are first identified by perfomung the partial distance search (PDS) in the wavelet domain. To implement the PDS in hardware, subspace search, bitplane reduction, multiple-coefficient accumulation and multiple-module computation techniques are employed for the effective reduction of the area complexity and computation latency. The proposed implementation has been embedded in a softcore CPU for physical performance measurement. Experimental results show that the implementation provides acosteffective solution to the FPGA realization of kNN classification systems where both high throughput and low area cost are desired.

AB - This paper presents a novel architecture for k-nearest neighbor (kNN) classification using field programmable gate array (FPGA). In the architecture, the first k closest vectors in the design set of a kNN classifier for each input vector are first identified by perfomung the partial distance search (PDS) in the wavelet domain. To implement the PDS in hardware, subspace search, bitplane reduction, multiple-coefficient accumulation and multiple-module computation techniques are employed for the effective reduction of the area complexity and computation latency. The proposed implementation has been embedded in a softcore CPU for physical performance measurement. Experimental results show that the implementation provides acosteffective solution to the FPGA realization of kNN classification systems where both high throughput and low area cost are desired.

KW - FPGA

KW - Knn

KW - Partial distance search

KW - Reconfigurable computing

UR - http://www.scopus.com/inward/record.url?scp=84855373576&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84855373576&partnerID=8YFLogxK

U2 - 10.1080/10798587.2011.10643159

DO - 10.1080/10798587.2011.10643159

M3 - Article

AN - SCOPUS:84855373576

VL - 17

SP - 431

EP - 446

JO - Intelligent Automation and Soft Computing

JF - Intelligent Automation and Soft Computing

SN - 1079-8587

IS - 4

ER -