Output details
11 - Computer Science and Informatics
Open University
Ameliorative missing value imputation for robust biological knowledge inference
<28>Based on the open-source Collateral Missing Value Imputation algorithm published in Bioinformatics (2005), which is an accepted benchmark imputation model, this paper quantitatively evaluates the downstream impact of missing values in microarray gene expression data. It crucially considers the biological significance of selected genes, with statistical analysis investigating the influence of missing values on subsequent biological analysis. The new imputation algorithm has been applied by a group at the Victorian Bioinformatics Consortium and also adapted to accurately impute missing data in wireless sensor networks.