Skip to main content
Figure 3 | BMC Bioinformatics

Figure 3

From: Missing value imputation for microarray gene expression data using histone acetylation information

Figure 3

Pearson correlation between imputed dataset and original complete dataset. Comparisons of Pearson correlation coefficients for KNN, LLS, knnHAI and llsHAI in six datasets under random model of missing values. The horizontal axis is the varying range of missing percentages from 1% to 20%. The vertical axis is Pearson correlation coefficients of 100 independent and random test runs for each method. We observed that the results are consistent with the NRMSE performances in Figure 1 across the whole test missing percentages. The genes imputed by llsHAI are most correlated with the original complete genes while the genes imputed by knnHAI are more correlated with the original complete genes than those imputed by KNN.

Back to article page