Validation of a Novel Traditional Chinese Medicine Pulse Diagnostic Model Using an Artificial Neural Network

Table 4

Comparison of the specificity, sensitivity, and accuracy of the best results of the models using different ANN training algorithms and the logistic regression.

Algorithm	Specificity; (%)	Sensitivity; (%)	Accuracy; (%)	Remarks
Bayesian regularization	73.46	84.49	79.27	15 hidden neurons
Levenberg-Marquardt	68.29	86.75	78.06	10 hidden neurons
Resilient backpropagation	72.19	83.91	78.41	10 hidden neurons
Probabilistic neural network	68.49	78.32	73.76	—