Performance comparison between Logistic regression, decision trees, and multilayer perceptron in predicting peripheral neuropathy in type 2 diabetes mellitus

• Published in: Chinese medical journal Vol. 125; no. 5; pp. 851 - 857
• Main Authors: Li, Chang-ping, Zhi, Xin-yue, Ma, Jun, Cui, Zhuang, Zhu, Zi-long, Zhang, Cui, Hu, Liang-ping
• Format: Journal Article
• Language: English
• Published: China Department of Health Statistics, College of Public Health,Tianjin Medical University, Tianjin 300070, China%Department of Epidemiology, College of Public Health,Tianjin Medical University, Tianjin 300070, China%Department of Internal Neurology, Tianjin Huanhu Hospital,Tianjin 300060, China%Key Laboratory of Advanced Energy Materials Chemistry(Ministry of Education), Nankai University, Tianjin 300071, China%Consulting Center of Biomedical Statistics, Academy of Military Medical Science, Beiiing 100850, China 01.03.2012
• Subjects: 2型糖尿病; Case-Control Studies; Decision Trees; Diabetes Mellitus, Type 2 - complications; Diabetic Neuropathies - diagnosis; Diabetic Neuropathies - etiology; Humans; Logistic Models; Logistic回归; 决策树; 周围神经; 多层感知器; 性能比较; 病变; 预测模型; multilayer perceptron; Logistic regression; diabetic peripheral neuropathy; decision tree
• ISSN: 0366-6999; 2542-5641
• DOI: 10.3760/cma.j.issn.0366-6999.2012.05.022

Background Various methods can be applied to build predictive models for the clinical data with binary outcome variable. This research aims to explore the process of constructing common predictive models, Logistic regression （LR）, decision tree （DT） and multilayer perceptron （MLP）, as well as focus on specific details when applying the methods mentioned above： what preconditions should be satisfied, how to set parameters of the model, how to screen variables and build accuracy models quickly and efficiently, and how to assess the generalization ability （that is, prediction performance） reliably by Monte Carlo method in the case of small sample size.