论文部分内容阅读
由于先前的研究工作不够综合和精确,不足于建立准确的高血压风险评估系统,根据2231个正常样本及823个高血压样本计算的信息增益,对高血压致病因素的重要程度进行了排序,总共建立和测试了42个不同的神经网络模型,发现了一个输入为26个致病因素的神经网络模型,其预测精度远高于先前研究取得的81.61%,该模型关于“是否高血压”、“收缩压”、“舒张压”的预测符合率分别为95.79%,98.22%和98.41%.基于发现的神经网络模型及面向对象的技术,开发了一个能自动收集新样本、学习新样本并能改进预测精度的高血压风险在线评估系统.
Because previous studies were not comprehensive and accurate enough to establish an accurate risk assessment system for hypertension, the importance of hypertension-related risk factors was ranked according to the information gain calculated from 2231 normal samples and 823 hypertension samples. A total of 42 different neural network models were established and tested, and a neural network model with 26 causative factors was found. The prediction accuracy was much higher than 81.61% obtained from the previous study. The model was about whether hypertension, The predicted coincidence rates were 95.79%, 98.22% and 98.41% respectively, based on the neural network models and object-oriented techniques.A new method was developed to automatically collect new samples , An online assessment of hypertension risk for learning new samples and improving prediction accuracy.