Service: Machine Learning in Editing and Imputation - Automated editing and correcting of data

One important step in statistical data processing is editing and imputation of missing or implausible observations. The Federal Statistical Office Germany is developing approaches to allow for automated editing and imputation based on observed values and plausibility rules. For the task machine learning methods are one important tool. The goal of the automated editing and imputation projects is to relief staff from editing and imputation tasks and to improve timeliness of statistical results at the same time. Often automated approaches are imperative, because digitization leads to larger data sets that need to be processed in a timely manner.