论文部分内容阅读
用Q型逐次信息群分对白银矿区42名矽肺患者和41名正常人头发样的元素谱Cr、Zn、Mg、Al、Cd进行无监督模式识别,获得分类清晰的谱系图,83个样本的判别正确率达98.8%。这一结果表明,元素谱的Q型逐次信息群分可望成为研究和预测矽肺病的一种新技术。
By using Q-type sequential information cluster analysis, 42 samples of silicosis patients and 41 normal hair-like elements of silver, Cr, Zn, Mg, Al and Cd in silver mine area were identified by unsupervised pattern recognition. Discrimination correct rate of 98.8%. This result indicates that the Q-type sequential information component of the elemental spectrum is expected to be a new technique for studying and predicting silicosis.