Multi-Sensor Activity Monitoring: Combination of Models with Class-Specific Voting
This paper presents a multi-sensor model combination system with class-specific voting for physical activity monitoring, which combines multiple classifiers obtained by splicing sensor data from different nodes into new data frames to improve the diversity of model inputs. Data obtained from a weara...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2019-06-01
|
| Series: | Information |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2078-2489/10/6/197 |
| _version_ | 1818245764817616896 |
|---|---|
| author | Lingfei Mo Lujie Zeng Shaopeng Liu Robert X. Gao |
| author_facet | Lingfei Mo Lujie Zeng Shaopeng Liu Robert X. Gao |
| author_sort | Lingfei Mo |
| collection | DOAJ |
| description | This paper presents a multi-sensor model combination system with class-specific voting for physical activity monitoring, which combines multiple classifiers obtained by splicing sensor data from different nodes into new data frames to improve the diversity of model inputs. Data obtained from a wearable multi-sensor wireless integrated measurement system (WIMS) consisting of two accelerometers and one ventilation sensor have been analysed to identify 10 different activity types of varying intensities performed by 110 voluntary participants. It is noted that each classifier shows better performance on some specific activity classes. Through class-specific weighted majority voting, the recognition accuracy of 10 PA types has been improved from 86% to 92% compared with the non-combination approach. Furthermore, the combination method has shown to be effective in reducing the subject-to-subject variability (standard deviation of recognition accuracies across subjects) in activity recognition and has better performance in monitoring physical activities of varying intensities than traditional homogeneous classifiers. |
| first_indexed | 2024-12-12T14:38:06Z |
| format | Article |
| id | doaj.art-1d6df497ff5443dba932309ffd41e427 |
| institution | Directory Open Access Journal |
| issn | 2078-2489 |
| language | English |
| last_indexed | 2024-12-12T14:38:06Z |
| publishDate | 2019-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Information |
| spelling | doaj.art-1d6df497ff5443dba932309ffd41e4272022-12-22T00:21:19ZengMDPI AGInformation2078-24892019-06-0110619710.3390/info10060197info10060197Multi-Sensor Activity Monitoring: Combination of Models with Class-Specific VotingLingfei Mo0Lujie Zeng1Shaopeng Liu2Robert X. Gao3School of Instrument Science and Engineer, Southeast University, Nanjing 210096, ChinaSchool of Instrument Science and Engineer, Southeast University, Nanjing 210096, ChinaGE Global Research, Niskayuna, NY 12309, USADepartment of Mechanical and Aerospace Engineering, Case Western Reserve University, Cleveland, OH 44106, USAThis paper presents a multi-sensor model combination system with class-specific voting for physical activity monitoring, which combines multiple classifiers obtained by splicing sensor data from different nodes into new data frames to improve the diversity of model inputs. Data obtained from a wearable multi-sensor wireless integrated measurement system (WIMS) consisting of two accelerometers and one ventilation sensor have been analysed to identify 10 different activity types of varying intensities performed by 110 voluntary participants. It is noted that each classifier shows better performance on some specific activity classes. Through class-specific weighted majority voting, the recognition accuracy of 10 PA types has been improved from 86% to 92% compared with the non-combination approach. Furthermore, the combination method has shown to be effective in reducing the subject-to-subject variability (standard deviation of recognition accuracies across subjects) in activity recognition and has better performance in monitoring physical activities of varying intensities than traditional homogeneous classifiers.https://www.mdpi.com/2078-2489/10/6/197activity monitoringsupport vector machinemulti-sensor combinationweighted voting |
| spellingShingle | Lingfei Mo Lujie Zeng Shaopeng Liu Robert X. Gao Multi-Sensor Activity Monitoring: Combination of Models with Class-Specific Voting Information activity monitoring support vector machine multi-sensor combination weighted voting |
| title | Multi-Sensor Activity Monitoring: Combination of Models with Class-Specific Voting |
| title_full | Multi-Sensor Activity Monitoring: Combination of Models with Class-Specific Voting |
| title_fullStr | Multi-Sensor Activity Monitoring: Combination of Models with Class-Specific Voting |
| title_full_unstemmed | Multi-Sensor Activity Monitoring: Combination of Models with Class-Specific Voting |
| title_short | Multi-Sensor Activity Monitoring: Combination of Models with Class-Specific Voting |
| title_sort | multi sensor activity monitoring combination of models with class specific voting |
| topic | activity monitoring support vector machine multi-sensor combination weighted voting |
| url | https://www.mdpi.com/2078-2489/10/6/197 |
| work_keys_str_mv | AT lingfeimo multisensoractivitymonitoringcombinationofmodelswithclassspecificvoting AT lujiezeng multisensoractivitymonitoringcombinationofmodelswithclassspecificvoting AT shaopengliu multisensoractivitymonitoringcombinationofmodelswithclassspecificvoting AT robertxgao multisensoractivitymonitoringcombinationofmodelswithclassspecificvoting |