Stacked Autoencoder and Meta-Learning based Heterogeneous Domain Adaptation for Human Activity Recognition
The field of human activity recognition (HAR) using machine learning approaches has gained a lot of interest in the research community due to its empowerment of automation and autonomous systems in industries and homes with respect to the given context and due to the increasing number of smart weara...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
University of Sindh
2022-07-01
|
| Series: | University of Sindh Journal of Information and Communication Technology |
| Subjects: | |
| Online Access: | https://sujo.usindh.edu.pk/index.php/USJICT/article/view/6267 |
| _version_ | 1827927606518874112 |
|---|---|
| author | Sajjad Ali Awan Sunder Ali Khowaja Kamran Taj Pathan |
| author_facet | Sajjad Ali Awan Sunder Ali Khowaja Kamran Taj Pathan |
| author_sort | Sajjad Ali Awan |
| collection | DOAJ |
| description | The field of human activity recognition (HAR) using machine learning approaches has gained a lot of interest in the research community due to its empowerment of automation and autonomous systems in industries and homes with respect to the given context and due to the increasing number of smart wearable devices. However, it is challenging to achieve a considerable accuracy for recognizing actions with diverse set of wearable devices due to their variance in feature spaces, sampling rate, units, sensor modalities and so forth. Furthermore, collecting annotated data has always been a serious issue in the machine learning community. Domain adaptation is a field that helps to cope with the issue by training on the source domain and labeling the samples in the target domain, however, due to the aforementioned variances (heterogeneity) in wearable sensor data, the action recognition accuracy remains on the lower side. Existing studies try to make the target domain feature space compliant with the source domain to improve the results, but it assumes that the system has a prior knowledge of the feature space of the target domain, which does not reflect real-world implication. In this regard, we propose stacked autoencoder and meta-learning based heterogeneous domain adaptation (SAM- HDD) network. The stacked autoencoder part is trained on the source domain feature space to extract the latent representation and train the employed classifiers, accordingly. The classification probabilities from the classifiers are trained with meta-learner to further improve the recognition performance. The data from tar- get domain undergoes the encoding layers of the trained stacked autoencoders to extract the latent representations, followed by the classification of label from the trained classifiers and meta- learner. The results show that the proposed approach is efficient in terms of accuracy score and achieves best results among the existing works, respectively |
| first_indexed | 2024-03-13T05:55:41Z |
| format | Article |
| id | doaj.art-2c1d4c32ab84472bab1198c78be3a3d2 |
| institution | Directory Open Access Journal |
| issn | 2521-5582 2523-1235 |
| language | English |
| last_indexed | 2024-03-13T05:55:41Z |
| publishDate | 2022-07-01 |
| publisher | University of Sindh |
| record_format | Article |
| series | University of Sindh Journal of Information and Communication Technology |
| spelling | doaj.art-2c1d4c32ab84472bab1198c78be3a3d22023-06-13T06:04:24ZengUniversity of SindhUniversity of Sindh Journal of Information and Communication Technology2521-55822523-12352022-07-016231376267Stacked Autoencoder and Meta-Learning based Heterogeneous Domain Adaptation for Human Activity RecognitionSajjad Ali Awan0Sunder Ali Khowaja1Kamran Taj Pathan2Dr. A.H.S Bukhari Postgraduate Center of ICT, University of Sindh, Jamshoro.Department of Telecommunication Engineering, Faculty of Engineering and Technology, University of Sindh, JamshoroDepartment of Software Engineering, Faculty of Engineering and Technology, University of Sindh, JamshoroThe field of human activity recognition (HAR) using machine learning approaches has gained a lot of interest in the research community due to its empowerment of automation and autonomous systems in industries and homes with respect to the given context and due to the increasing number of smart wearable devices. However, it is challenging to achieve a considerable accuracy for recognizing actions with diverse set of wearable devices due to their variance in feature spaces, sampling rate, units, sensor modalities and so forth. Furthermore, collecting annotated data has always been a serious issue in the machine learning community. Domain adaptation is a field that helps to cope with the issue by training on the source domain and labeling the samples in the target domain, however, due to the aforementioned variances (heterogeneity) in wearable sensor data, the action recognition accuracy remains on the lower side. Existing studies try to make the target domain feature space compliant with the source domain to improve the results, but it assumes that the system has a prior knowledge of the feature space of the target domain, which does not reflect real-world implication. In this regard, we propose stacked autoencoder and meta-learning based heterogeneous domain adaptation (SAM- HDD) network. The stacked autoencoder part is trained on the source domain feature space to extract the latent representation and train the employed classifiers, accordingly. The classification probabilities from the classifiers are trained with meta-learner to further improve the recognition performance. The data from tar- get domain undergoes the encoding layers of the trained stacked autoencoders to extract the latent representations, followed by the classification of label from the trained classifiers and meta- learner. The results show that the proposed approach is efficient in terms of accuracy score and achieves best results among the existing works, respectivelyhttps://sujo.usindh.edu.pk/index.php/USJICT/article/view/6267human activity recognitiondomain adaptationheterogeneous domain adaptationstacked autoencodermeta-learning |
| spellingShingle | Sajjad Ali Awan Sunder Ali Khowaja Kamran Taj Pathan Stacked Autoencoder and Meta-Learning based Heterogeneous Domain Adaptation for Human Activity Recognition University of Sindh Journal of Information and Communication Technology human activity recognition domain adaptation heterogeneous domain adaptation stacked autoencoder meta-learning |
| title | Stacked Autoencoder and Meta-Learning based Heterogeneous Domain Adaptation for Human Activity Recognition |
| title_full | Stacked Autoencoder and Meta-Learning based Heterogeneous Domain Adaptation for Human Activity Recognition |
| title_fullStr | Stacked Autoencoder and Meta-Learning based Heterogeneous Domain Adaptation for Human Activity Recognition |
| title_full_unstemmed | Stacked Autoencoder and Meta-Learning based Heterogeneous Domain Adaptation for Human Activity Recognition |
| title_short | Stacked Autoencoder and Meta-Learning based Heterogeneous Domain Adaptation for Human Activity Recognition |
| title_sort | stacked autoencoder and meta learning based heterogeneous domain adaptation for human activity recognition |
| topic | human activity recognition domain adaptation heterogeneous domain adaptation stacked autoencoder meta-learning |
| url | https://sujo.usindh.edu.pk/index.php/USJICT/article/view/6267 |
| work_keys_str_mv | AT sajjadaliawan stackedautoencoderandmetalearningbasedheterogeneousdomainadaptationforhumanactivityrecognition AT sunderalikhowaja stackedautoencoderandmetalearningbasedheterogeneousdomainadaptationforhumanactivityrecognition AT kamrantajpathan stackedautoencoderandmetalearningbasedheterogeneousdomainadaptationforhumanactivityrecognition |