论文部分内容阅读
针对现有的 Bayesian网络学习方法都不能有效处理缺失数据问题 ,论文给出了两种处理不完整数据问题的方法 :一种方法是先把不完整的数据集修复成完整的数据集 ,利用完整的数据集进行计算 ,并将结果作为不完整数据集对应情况的近似 ;另一种方法是直接使用不完整的数据集进行近似计算 ,而这种近似计算是渐进正确的。实验结果表明前一种方法计算结果准确 ,但效率较低 ;后一种方法效率较高 ,在数据量比较大时能达到很好的效果 ;而且这两种方法的性能比其它处理缺失数据的方法效果要好。
For the existing Bayesian network learning methods can not effectively deal with the problem of missing data, the paper gives two methods to deal with the problem of incomplete data: One method is to repair the incomplete data set into a complete data set, , And the result is taken as an approximation of the corresponding situation of the incomplete data set. The other method is to use the incomplete data set directly for approximate calculation, and the approximate calculation is asymptotic correct. The experimental results show that the former method is accurate, but less efficient. The latter method is more efficient and achieves good results when the amount of data is large. Moreover, the performance of the two methods is better than that of the other methods Method is better.