with the kernel weight based on similarly between target gene and sample genes, which localize missing value estimation, a new method based on weighted regression is presented.
in order to exclude the effect of dissimilar genes, a new missing value estimation method based on knn-svr is proposed.
在不增加实验次数的情况下,缺失值估计是降低缺失数据对后续分析影响的有效方法。
secondly, the dataset was transformed to a transaction database(tdb) through numbering all the effective data, eliminating the missing value and adding transaction identification for each point.
there is missing value in microarray experiments and it will affect the stability and precision of the expression data analysis.
在基因芯片实验中,数据缺失客观存在,并在一定程度上影响芯片数据后续分析结果的準确性。
results the multiply imputed data generated by mi procedure represent the uncertainty about the missing value and help us use data available to estimate population parameter more effectively.