Paper Info
Title
Diagnosis Of Complications Of Type 2 Diabetes Based On Weighted Multi-Label Small Sphere And Large Margin Machine
Abstract
At present, type 2 diabetes mellitus (T2DM) is one of the most serious and critical health problems. Persistent hyperglycemia of diabetic patients can lead to other complications, such as macrovascular, microvascular, neuropathy, which are the main cause of death in diabetic patients. Therefore, it is an urgent task to diagnose the complications. To address the above issue, we turn it into a multi-label classification problem by taking macrovascular, microvascular, neuropathy as three labels. Furthermore, we find that it is an imbalanced classification problem for each label. Thus, a novel weighted multi-label small sphere and large margin machine (WML-SSLM) is proposed to diagnose the complications from T2DM in this paper, which is constructed by introducing the binary relevance (BR) method to SSLM. Compared with the BR method, WML-SSLM considers the relevance of labels by giving different weights for different instances. Taking the diabetes dataset from the Chinese PLA General Hospital as the research object, the diagnosis of the macrovascular, microvascular, and neuropathy from T2DM are studied by using our proposed WML-SSLM. The experimental results show that WML-SSLM can effectively deal with the prediction of complications of T2DM. Besides, the relevant features of each complication are analyzed by using the student's t-test.
Year
DOI
Venue
2021
10.1007/s10489-020-01824-y
APPLIED INTELLIGENCE
Keywords
DocType
Volume
Type 2 diabetes, Complications, Hyper-sphere support vector machine, Multi-label learning
Journal
51
Issue
ISSN
Citations 
1
0924-669X
0
PageRank 
References 
Authors
0.34
0
4
Name
Order
Citations
PageRank
Hongmei Wang13113.44
Yitian Xu248935.06
Qian Chen34511.68
Xinye Wang400.34