Incremental-precision based feature computation and multi-level classification for low-energy internet-of-things

Sandhya Koteshwara, Keshab K Parhi

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

This paper presents a novel technique to reduce energy consumption of a machine learning classifier based on incremental-precision feature computation and classification. Specifically, the algorithm starts with features computed using the lowest possible precision. Depending on the classification accuracy, the features of the previous level are combined with features of the incremental-precision to compute the features in higher-precision. This process is continued till a desired accuracy is obtained. A certain threshold that allows many samples to be classified using a low-precision classifier can reduce energy consumption, but increases misclassification error. To implement hardware which provides the required updates in precision, an incremental-precision architecture based on data-path decomposition is proposed. One novel aspect of this work lies in the design of appropriate thresholds for multi-level classification using training data such that a family of designs can be obtained that enable trade-offs between classification accuracy and energy consumption. Another novel aspect involves the design of hardware architectures based on data-path decomposition which can incrementally increase precision upon demand. Using a seizure detection example, it is shown that the proposed incremental-precision-based multi-level classification approach can reduce energy consumption by 35% while maintaining high sensitivity, or by about 50% at the expense of 15% degradation in sensitivity compared with similar approaches to seizure detection in literature. The reduction in energy is achieved at the expense of small area, timing and memory overheads as multiple classification steps are used instead of a single step.

Original languageEnglish (US)
Article number8358752
Pages (from-to)822-835
Number of pages14
JournalIEEE Journal on Emerging and Selected Topics in Circuits and Systems
Volume8
Issue number4
DOIs
StatePublished - Dec 1 2018

Fingerprint

Energy utilization
Classifiers
Decomposition
Hardware
Learning systems
Internet of things
Data storage equipment
Degradation

Keywords

  • Approximate computing
  • Datapath decomposition
  • Energy reduction
  • Incremental-precision computation
  • Machine learning
  • Multi-level classification
  • Seizure detection

Cite this

@article{53b1732569fc4ca7aecb713d7c3fdf6d,
title = "Incremental-precision based feature computation and multi-level classification for low-energy internet-of-things",
abstract = "This paper presents a novel technique to reduce energy consumption of a machine learning classifier based on incremental-precision feature computation and classification. Specifically, the algorithm starts with features computed using the lowest possible precision. Depending on the classification accuracy, the features of the previous level are combined with features of the incremental-precision to compute the features in higher-precision. This process is continued till a desired accuracy is obtained. A certain threshold that allows many samples to be classified using a low-precision classifier can reduce energy consumption, but increases misclassification error. To implement hardware which provides the required updates in precision, an incremental-precision architecture based on data-path decomposition is proposed. One novel aspect of this work lies in the design of appropriate thresholds for multi-level classification using training data such that a family of designs can be obtained that enable trade-offs between classification accuracy and energy consumption. Another novel aspect involves the design of hardware architectures based on data-path decomposition which can incrementally increase precision upon demand. Using a seizure detection example, it is shown that the proposed incremental-precision-based multi-level classification approach can reduce energy consumption by 35{\%} while maintaining high sensitivity, or by about 50{\%} at the expense of 15{\%} degradation in sensitivity compared with similar approaches to seizure detection in literature. The reduction in energy is achieved at the expense of small area, timing and memory overheads as multiple classification steps are used instead of a single step.",
keywords = "Approximate computing, Datapath decomposition, Energy reduction, Incremental-precision computation, Machine learning, Multi-level classification, Seizure detection",
author = "Sandhya Koteshwara and Parhi, {Keshab K}",
year = "2018",
month = "12",
day = "1",
doi = "10.1109/JETCAS.2018.2836319",
language = "English (US)",
volume = "8",
pages = "822--835",
journal = "IEEE Journal on Emerging and Selected Topics in Circuits and Systems",
issn = "2156-3357",
publisher = "IEEE Circuits and Systems Society",
number = "4",

}

TY - JOUR

T1 - Incremental-precision based feature computation and multi-level classification for low-energy internet-of-things

AU - Koteshwara, Sandhya

AU - Parhi, Keshab K

PY - 2018/12/1

Y1 - 2018/12/1

N2 - This paper presents a novel technique to reduce energy consumption of a machine learning classifier based on incremental-precision feature computation and classification. Specifically, the algorithm starts with features computed using the lowest possible precision. Depending on the classification accuracy, the features of the previous level are combined with features of the incremental-precision to compute the features in higher-precision. This process is continued till a desired accuracy is obtained. A certain threshold that allows many samples to be classified using a low-precision classifier can reduce energy consumption, but increases misclassification error. To implement hardware which provides the required updates in precision, an incremental-precision architecture based on data-path decomposition is proposed. One novel aspect of this work lies in the design of appropriate thresholds for multi-level classification using training data such that a family of designs can be obtained that enable trade-offs between classification accuracy and energy consumption. Another novel aspect involves the design of hardware architectures based on data-path decomposition which can incrementally increase precision upon demand. Using a seizure detection example, it is shown that the proposed incremental-precision-based multi-level classification approach can reduce energy consumption by 35% while maintaining high sensitivity, or by about 50% at the expense of 15% degradation in sensitivity compared with similar approaches to seizure detection in literature. The reduction in energy is achieved at the expense of small area, timing and memory overheads as multiple classification steps are used instead of a single step.

AB - This paper presents a novel technique to reduce energy consumption of a machine learning classifier based on incremental-precision feature computation and classification. Specifically, the algorithm starts with features computed using the lowest possible precision. Depending on the classification accuracy, the features of the previous level are combined with features of the incremental-precision to compute the features in higher-precision. This process is continued till a desired accuracy is obtained. A certain threshold that allows many samples to be classified using a low-precision classifier can reduce energy consumption, but increases misclassification error. To implement hardware which provides the required updates in precision, an incremental-precision architecture based on data-path decomposition is proposed. One novel aspect of this work lies in the design of appropriate thresholds for multi-level classification using training data such that a family of designs can be obtained that enable trade-offs between classification accuracy and energy consumption. Another novel aspect involves the design of hardware architectures based on data-path decomposition which can incrementally increase precision upon demand. Using a seizure detection example, it is shown that the proposed incremental-precision-based multi-level classification approach can reduce energy consumption by 35% while maintaining high sensitivity, or by about 50% at the expense of 15% degradation in sensitivity compared with similar approaches to seizure detection in literature. The reduction in energy is achieved at the expense of small area, timing and memory overheads as multiple classification steps are used instead of a single step.

KW - Approximate computing

KW - Datapath decomposition

KW - Energy reduction

KW - Incremental-precision computation

KW - Machine learning

KW - Multi-level classification

KW - Seizure detection

UR - http://www.scopus.com/inward/record.url?scp=85046816380&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046816380&partnerID=8YFLogxK

U2 - 10.1109/JETCAS.2018.2836319

DO - 10.1109/JETCAS.2018.2836319

M3 - Article

VL - 8

SP - 822

EP - 835

JO - IEEE Journal on Emerging and Selected Topics in Circuits and Systems

JF - IEEE Journal on Emerging and Selected Topics in Circuits and Systems

SN - 2156-3357

IS - 4

M1 - 8358752

ER -