论文部分内容阅读
将BP神经网络方法用于血吸虫性肝硬化的临床中疾病诊断研究。该方法首先利用主成分分析法提取特征变量的主成分，作为BP神经网络的输入；然后选择样本对BP神经网络进行训练；最后用训练好的BP神经网络对慢性血吸虫性肝硬化进行判别。仿真结果表明，该方法能有效区分慢性血吸虫性肝硬化和非血吸虫性肝硬化，同时，利用主成分分析法不仅简化了BP网络模型，还提高了预测的精度。“,”The method of BP neurallnetwork was used in the clinicalldisease diagnosis of the schistosomiasis. Firstly, the principallcomponent of characteristic variables were extracted which were as the inputs of the BP neurallnetwork; secondly, the BP neurallnetwork was trained by the selected samples; finally, the schistosomiasis cirrhosis was recognized by the trained BP neurallnetwork. The simulation results show that the BP neurallnetwork can Effectively distinguish the schistosomiasis cirrhosis from the non-schistosomiasis cirrhosis, and the introducing of the method of principallcomponent analysis not only simplifies the BP neurallnetwork modellbut also improve the prediction accuracy.