This application will guide you through the process of eliminating data columns that are useless or even harmful to your analysis. The average error in % from a cross-validation procedure is used as a measure for the dataset quality. Cross-validation is tenfold and based here on a decision tree. Final decisions are recorded in an audit report and saved in the file, auditReport.xls. The whole workflow has been implemented to run interactively on the KNIME WebPortal. On the WebPortal click START to begin.