Homogeneous Data Normalization and Deep Learning: A Case Study in Human Activity Classification
One class of applications for human activity recognition methods is found in mobile devices for monitoring older adults and people with special needs. Recently, many studies were performed to create intelligent methods for the recognition of human activities. However, the different mobile devices in...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-11-01
|
Series: | Future Internet |
Subjects: | |
Online Access: | https://www.mdpi.com/1999-5903/12/11/194 |
id |
doaj-a8e9f2ffaafc436e94acf4aec2a85708 |
---|---|
record_format |
Article |
spelling |
doaj-a8e9f2ffaafc436e94acf4aec2a857082020-11-25T04:10:40ZengMDPI AGFuture Internet1999-59032020-11-011219419410.3390/fi12110194Homogeneous Data Normalization and Deep Learning: A Case Study in Human Activity ClassificationIvan Miguel Pires0Faisal Hussain1Nuno M. Garcia2Petre Lameski3Eftim Zdravevski4Instituto de Telecomunicações, Universidade da Beira Interior, 6200-001 Covilhã, PortugalDepartment of Computer Engineering, University of Engineering and Technology (UET), Taxila 47080, PakistanInstituto de Telecomunicações, Universidade da Beira Interior, 6200-001 Covilhã, PortugalFaculty of Computer Science and Engineering, University Ss Cyril and Methodius, 1000 Skopje, MacedoniaFaculty of Computer Science and Engineering, University Ss Cyril and Methodius, 1000 Skopje, MacedoniaOne class of applications for human activity recognition methods is found in mobile devices for monitoring older adults and people with special needs. Recently, many studies were performed to create intelligent methods for the recognition of human activities. However, the different mobile devices in the market acquire the data from sensors at different frequencies. This paper focuses on implementing four data normalization techniques, i.e., MaxAbsScaler, MinMaxScaler, RobustScaler, and Z-Score. Subsequently, we evaluate the impact of the normalization algorithms with deep neural networks (DNN) for the classification of the human activities. The impact of the data normalization was counterintuitive, resulting in a degradation of performance. Namely, when using the accelerometer data, the accuracy dropped from about 79% to only 53% for the best normalization approach. Similarly, for the gyroscope data, the accuracy without normalization was about 81.5%, whereas with the best normalization, it was only 60%. It can be concluded that data normalization techniques are not helpful in classification problems with homogeneous data.https://www.mdpi.com/1999-5903/12/11/194human activitiesdata normalizationdata classificationsensorsmobile devicesdata processing |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Ivan Miguel Pires Faisal Hussain Nuno M. Garcia Petre Lameski Eftim Zdravevski |
spellingShingle |
Ivan Miguel Pires Faisal Hussain Nuno M. Garcia Petre Lameski Eftim Zdravevski Homogeneous Data Normalization and Deep Learning: A Case Study in Human Activity Classification Future Internet human activities data normalization data classification sensors mobile devices data processing |
author_facet |
Ivan Miguel Pires Faisal Hussain Nuno M. Garcia Petre Lameski Eftim Zdravevski |
author_sort |
Ivan Miguel Pires |
title |
Homogeneous Data Normalization and Deep Learning: A Case Study in Human Activity Classification |
title_short |
Homogeneous Data Normalization and Deep Learning: A Case Study in Human Activity Classification |
title_full |
Homogeneous Data Normalization and Deep Learning: A Case Study in Human Activity Classification |
title_fullStr |
Homogeneous Data Normalization and Deep Learning: A Case Study in Human Activity Classification |
title_full_unstemmed |
Homogeneous Data Normalization and Deep Learning: A Case Study in Human Activity Classification |
title_sort |
homogeneous data normalization and deep learning: a case study in human activity classification |
publisher |
MDPI AG |
series |
Future Internet |
issn |
1999-5903 |
publishDate |
2020-11-01 |
description |
One class of applications for human activity recognition methods is found in mobile devices for monitoring older adults and people with special needs. Recently, many studies were performed to create intelligent methods for the recognition of human activities. However, the different mobile devices in the market acquire the data from sensors at different frequencies. This paper focuses on implementing four data normalization techniques, i.e., MaxAbsScaler, MinMaxScaler, RobustScaler, and Z-Score. Subsequently, we evaluate the impact of the normalization algorithms with deep neural networks (DNN) for the classification of the human activities. The impact of the data normalization was counterintuitive, resulting in a degradation of performance. Namely, when using the accelerometer data, the accuracy dropped from about 79% to only 53% for the best normalization approach. Similarly, for the gyroscope data, the accuracy without normalization was about 81.5%, whereas with the best normalization, it was only 60%. It can be concluded that data normalization techniques are not helpful in classification problems with homogeneous data. |
topic |
human activities data normalization data classification sensors mobile devices data processing |
url |
https://www.mdpi.com/1999-5903/12/11/194 |
work_keys_str_mv |
AT ivanmiguelpires homogeneousdatanormalizationanddeeplearningacasestudyinhumanactivityclassification AT faisalhussain homogeneousdatanormalizationanddeeplearningacasestudyinhumanactivityclassification AT nunomgarcia homogeneousdatanormalizationanddeeplearningacasestudyinhumanactivityclassification AT petrelameski homogeneousdatanormalizationanddeeplearningacasestudyinhumanactivityclassification AT eftimzdravevski homogeneousdatanormalizationanddeeplearningacasestudyinhumanactivityclassification |
_version_ |
1724419760873013248 |