Improve QoS for multi-body sensor analytics in smart healthcare system using machine learning algorithm
Embracing significant learning methods for human lead affirmation has shown suitable in taking out discriminants from the coarse information packs obtained from body-mounted sensors. But human headway is ideal coded in a movement of moderate models, the standard AI strategy is to finished certificat...
Uloženo v:
| Vydáno v: | Journal of interdisciplinary mathematics Ročník 26; číslo 3; s. 393 - 405 |
|---|---|
| Hlavní autoři: | , , , , , |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
2023
|
| ISSN: | 0972-0502, 2169-012X |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Shrnutí: | Embracing significant learning methods for human lead affirmation has shown suitable in taking out discriminants from the coarse information packs obtained from body-mounted sensors. But human headway is ideal coded in a movement of moderate models, the standard AI strategy is to finished certification obligations without taking advantage of the normal relationship between analysis information tests. This paper proposes the use of (DRNN) to manufacture a psychological model that can get critical distance conditions with factor-length input position. We present unidirectional, bidirectional, and comfortable models concerning DRNNs with LSTM and finding parameters using sporadic benchmark datasets. Exploratory results show that the proposed model is superior to a standard AI-based system. SVM and Nearest Neighbour Method (KNN). Moreover, In this Paper implementation smart system runs in tendency to other significant learning techniques like Deep Trust Organization (DBN) and CNN. Human Action Acknowledgment (HAR) assignments were consistently made using arranged highlights got by heuristic cycles. |
|---|---|
| ISSN: | 0972-0502 2169-012X |
| DOI: | 10.47974/JIM-1670 |