Activity recognition with android phone using mixture-of-experts co-trained with labeled and unlabeled data

Young Seol Lee, Sung Bae Cho

Research output: Contribution to journalArticle

48 Citations (Scopus)

Abstract

As the number of smartphone users has grown recently, many context-aware services have been studied and launched. Activity recognition becomes one of the important issues for user adaptive services on the mobile phones. Even though many researchers have attempted to recognize a user's activities on a mobile device, it is still difficult to infer human activities from uncertain, incomplete and insufficient mobile sensor data. We present a method to recognize a person's activities from sensors in a mobile phone using mixture-of-experts (ME) model. In order to train the ME model, we have applied global-local co-training (GLCT) algorithm with both labeled and unlabeled data to improve the performance. The GLCT is a variation of co-training that uses a global model and a local model together. To evaluate the usefulness of the proposed method, we have conducted experiments using real datasets collected from Google Android smartphones. This paper is a revised and extended version of a paper that was presented at HAIS 2011.

Original languageEnglish
Pages (from-to)106-115
Number of pages10
JournalNeurocomputing
Volume126
DOIs
Publication statusPublished - 2014 Feb 27

Fingerprint

Cell Phones
Smartphones
Mobile phones
Human Activities
Research Personnel
Equipment and Supplies
Sensors
Mobile devices
Smartphone
Experiments
Datasets

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Cognitive Neuroscience
  • Artificial Intelligence

Cite this

@article{d2a03ddd165241459433b964adc46a72,
title = "Activity recognition with android phone using mixture-of-experts co-trained with labeled and unlabeled data",
abstract = "As the number of smartphone users has grown recently, many context-aware services have been studied and launched. Activity recognition becomes one of the important issues for user adaptive services on the mobile phones. Even though many researchers have attempted to recognize a user's activities on a mobile device, it is still difficult to infer human activities from uncertain, incomplete and insufficient mobile sensor data. We present a method to recognize a person's activities from sensors in a mobile phone using mixture-of-experts (ME) model. In order to train the ME model, we have applied global-local co-training (GLCT) algorithm with both labeled and unlabeled data to improve the performance. The GLCT is a variation of co-training that uses a global model and a local model together. To evaluate the usefulness of the proposed method, we have conducted experiments using real datasets collected from Google Android smartphones. This paper is a revised and extended version of a paper that was presented at HAIS 2011.",
author = "Lee, {Young Seol} and Cho, {Sung Bae}",
year = "2014",
month = "2",
day = "27",
doi = "10.1016/j.neucom.2013.05.044",
language = "English",
volume = "126",
pages = "106--115",
journal = "Review of Economic Dynamics",
issn = "1094-2025",
publisher = "Academic Press Inc.",

}

Activity recognition with android phone using mixture-of-experts co-trained with labeled and unlabeled data. / Lee, Young Seol; Cho, Sung Bae.

In: Neurocomputing, Vol. 126, 27.02.2014, p. 106-115.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Activity recognition with android phone using mixture-of-experts co-trained with labeled and unlabeled data

AU - Lee, Young Seol

AU - Cho, Sung Bae

PY - 2014/2/27

Y1 - 2014/2/27

N2 - As the number of smartphone users has grown recently, many context-aware services have been studied and launched. Activity recognition becomes one of the important issues for user adaptive services on the mobile phones. Even though many researchers have attempted to recognize a user's activities on a mobile device, it is still difficult to infer human activities from uncertain, incomplete and insufficient mobile sensor data. We present a method to recognize a person's activities from sensors in a mobile phone using mixture-of-experts (ME) model. In order to train the ME model, we have applied global-local co-training (GLCT) algorithm with both labeled and unlabeled data to improve the performance. The GLCT is a variation of co-training that uses a global model and a local model together. To evaluate the usefulness of the proposed method, we have conducted experiments using real datasets collected from Google Android smartphones. This paper is a revised and extended version of a paper that was presented at HAIS 2011.

AB - As the number of smartphone users has grown recently, many context-aware services have been studied and launched. Activity recognition becomes one of the important issues for user adaptive services on the mobile phones. Even though many researchers have attempted to recognize a user's activities on a mobile device, it is still difficult to infer human activities from uncertain, incomplete and insufficient mobile sensor data. We present a method to recognize a person's activities from sensors in a mobile phone using mixture-of-experts (ME) model. In order to train the ME model, we have applied global-local co-training (GLCT) algorithm with both labeled and unlabeled data to improve the performance. The GLCT is a variation of co-training that uses a global model and a local model together. To evaluate the usefulness of the proposed method, we have conducted experiments using real datasets collected from Google Android smartphones. This paper is a revised and extended version of a paper that was presented at HAIS 2011.

UR - http://www.scopus.com/inward/record.url?scp=84887606604&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84887606604&partnerID=8YFLogxK

U2 - 10.1016/j.neucom.2013.05.044

DO - 10.1016/j.neucom.2013.05.044

M3 - Article

AN - SCOPUS:84887606604

VL - 126

SP - 106

EP - 115

JO - Review of Economic Dynamics

JF - Review of Economic Dynamics

SN - 1094-2025

ER -