The design of regional medical cloud computing information platform based on deep learning

In order to solve the imbalance of medical resources in different regions, a regional medical cloud computing information platform based on reactive algorithm is constructed. First, an application-oriented elastic scaling algorithm based on long short-term memory network and back propagation neural...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:International journal of system assurance engineering and management Ročník 12; číslo 4; s. 757 - 764
Hlavní autor: Zhang, Kaidong
Médium: Journal Article
Jazyk:angličtina
Vydáno: New Delhi Springer India 01.08.2021
Springer Nature B.V
Témata:
Algorithms
Artificial neural networks
Back propagation networks
Cloud computing
Data mining
Deep learning
Engineering
Engineering Economics
Logistics
Machine learning
Marketing
Neural networks
Optimization
Organization
Original Article
Quality Control
Regional development
Regression models
Reliability
Response time
Safety and Risk
Visual programming languages
Cloud computing
Regional medical
Data mining
Reactive algorithm
ISSN:0975-6809, 0976-4348
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:In order to solve the imbalance of medical resources in different regions, a regional medical cloud computing information platform based on reactive algorithm is constructed. First, an application-oriented elastic scaling algorithm based on long short-term memory network and back propagation neural network (BPNN) is proposed. Then, based on cloud computing, a medical cloud data mining platform using Hadoop ecosystem is proposed. Finally, Visual Studio is used to develop regional medical cloud computing information platform, and the performance of the platform is tested. The experimental results show that the improved neural network algorithm has a loss and MAPE (mean absolute percentage error) value of 930 and 0.00031, respectively in the actual workload prediction, which is better than the algorithm before optimization. Moreover, it has the best fitting effect with the actual curve in the prediction of response time. In the strategy scheduling experiment, the loss of the model is 1.40222, the MAPE value is 0.34021, and the convergence time is 23 s, which is better than the test results of the model based on linear regression and BPNN. The experimental results suggest that the regional medical cloud computing information platform can solve the problem of unfair regional medical resources in the medical field to a certain extent.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0975-6809
0976-4348
DOI:10.1007/s13198-021-01075-1