The aim of the project is to develop a novel validation strategy of cardiac arrhythmia classification algorithms based on multiparametric data analysis of electrocardiography (ECG) data through metrological research. A novel synthetic reference database will be developed that will enable to investigate, for the first time, the uncertainty of modern data analysis approaches, such as machine learning in medicine and to contribute to standardising machine learning methods in health applications, specifically by establishing a novel metrological validation platform of such algorithms manifesting a digital traceability chain.