1 INTRODUCTION
Data Science (DS), and in particular Machine Learning (ML), systems are increasingly becoming a used instrument, applied to all application domains and affecting our real life. Such systems can be defined as a set of one or more pipelines (or workflows), which take as input raw (unprocessed) data and returns actionable answers to questions in the form of machine learning models. In this paper, we focus on DS pipelines that leverage on ML, that we call ML pipelines.