为提高名词性属性实例差异的识别精度,优化分类算法性能,综合考虑实例的属性和类别特征,提出了一种基于条件概率分布的混合距离度量方法.首先,计算属性间以及属性与类别间条件概率分布的差异;其次,利用互信息对2种差异进行加权组合,得...为提高名词性属性实例差异的识别精度,优化分类算法性能,综合考虑实例的属性和类别特征,提出了一种基于条件概率分布的混合距离度量方法.首先,计算属性间以及属性与类别间条件概率分布的差异;其次,利用互信息对2种差异进行加权组合,得到新的混合距离度量;最后,利用K-近邻算法在20个UCI(University of California Irvine)数据集上进行仿真实验,并将其应用于儿童阑尾炎的诊断和治疗.结果表明:较重叠度量等3种度量方法,本文提出的距离度量方法,显著提高了分类算法的准确率,具有较好的应用前景.展开更多
Some research work has been done on the long-term joint distribution of one tenth of the large wave heights with average periods. On the basis of research, the conditional probability distribution of average periods t...Some research work has been done on the long-term joint distribution of one tenth of the large wave heights with average periods. On the basis of research, the conditional probability distribution of average periods to the wave heights are derived, and the range and the mode of average wave period to return wave height can be calculated easily. By the comparison of the results calculated by the method presented with those calculated by the Design Specification of Harbor Engineering of China, the method presented has more advantages than those of the Specification.展开更多
文摘为提高名词性属性实例差异的识别精度,优化分类算法性能,综合考虑实例的属性和类别特征,提出了一种基于条件概率分布的混合距离度量方法.首先,计算属性间以及属性与类别间条件概率分布的差异;其次,利用互信息对2种差异进行加权组合,得到新的混合距离度量;最后,利用K-近邻算法在20个UCI(University of California Irvine)数据集上进行仿真实验,并将其应用于儿童阑尾炎的诊断和治疗.结果表明:较重叠度量等3种度量方法,本文提出的距离度量方法,显著提高了分类算法的准确率,具有较好的应用前景.
文摘Some research work has been done on the long-term joint distribution of one tenth of the large wave heights with average periods. On the basis of research, the conditional probability distribution of average periods to the wave heights are derived, and the range and the mode of average wave period to return wave height can be calculated easily. By the comparison of the results calculated by the method presented with those calculated by the Design Specification of Harbor Engineering of China, the method presented has more advantages than those of the Specification.
