Table 2.

Machine learning method for predicting diabetic foot

Author	Learning Methods Purpose Prediction Result	Purpose	Prediction result
Zhang J, (2023) [28]	Linear regression	It is possible to calculate the regression coefficients and significance levels of various factors, and identify the independent risk factors for diabetic foot	The duration of diabetes, glycemic control, peripheral neuropathy, and foot deformities are independent risk factors for diabetic foot
Rachita Nanda, (2022)[29]	Random Forest	It can handle a large number of input variables and assess the importance of each variable, enabling comprehensive multi-factor analysis for predicting the risk of diabetic foot	Variables such as blood glucose levels, duration of diabetes, and foot examination results are considered important in predicting the risk of diabetic foot
Shiqi Wang, (2022) [30]	GBDT	The prediction accuracy can be improved by continuously optimizing model parameters, and it can handle nonlinear relationships and interactions	The prediction results from multiple decision trees are accumulated to generate a prediction probability for the risk of diabetic foot occurrence. The higher the probability, the greater the risk
Yu-Long Chen, (2023) [31]	Neural Networks	It can handle large-scale datasets and complex nonlinear relationships, capturing the intricate interactions among multiple risk factors	Input feature data is processed through a nonlinear transformation to obtain a prediction probability value. The higher the probability, the greater the risk
Shichai Hon, (2024) [32]	SVM	Analyze patient data to distinguish between infected and non-infected groups	By analyzing the data, patients are classified into infected and non infected groups
Xiaojin Zhang, (2022)[33]	XGBoost	It can automatically learn and capture complex patterns and features in patient data, employing various optimization techniques to improve prediction accuracy	The input feature data produces an output prediction probability value, where a higher number indicates a greater risk
Zhiyan Fu, (2024) [34]	Ensemble learning method	Combine the prediction results of multiple base learners to improve overall prediction performance	The prediction results from multiple base models are integrated, and the final result is presented in the form of a probability, indicating the risk level of a patient developing diabetic foot