We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
Bi-directional Long-term Short-term Memory Classification Method of Diabetes Datasets.
- Authors
Xu-Dong Li; Jie-Sheng Wang; Wen-Kuo Hao; Xiao-Rui Zhao
- Abstract
As an underlying disease, diabetes has become more and more common worldwide. The patients presented with persistent and long-term hyperglycemia. If the cause of diabetes is not identified and not treated in time, many complications will occur. Because the identification process is complicated, patients choose a medical center to visit a diagnostic center to consult a doctor. This study will focus on how to design a classification model to accurately determine whether a patient has diabetes. Therefore, this paper combines the trigonometric function pedigree arithmetic optimization algorithm (AOA) with the bi-directional long short-term memory (Bi-LSTM) neural network Based on two kinds of neural networks, 10 diabetes classification methods based on heuristic algorithm training network parameters are designed. Simulation experimental results demonstrate that the trigonometric function pedigree arithmetic optimization algorithm to improve the parameters of the Bi-LSTM network can significantly improve the classification accuracy. Finally, the Bi-LSTM classification model based on tanAOA has the best classification utility, and the classification accuracy is better than other classification methods.
- Subjects
SHORT-term memory; LONG-term memory; ARITHMETIC functions; TRIGONOMETRIC functions; ETIOLOGY of diabetes; HEURISTIC algorithms; CLASSIFICATION algorithms
- Publication
Engineering Letters, 2022, Vol 30, Issue 4, p1603
- ISSN
1816-093X
- Publication type
Article