In this talk we provide an overview of the most common tasks nowadays exploited by data scientists namely, classification and regression from the software engineer perspective.
In doing so, we provide a theoretical introduction concerning the modelling, purpose and limitations of the aforementioned tasks, along with a brief discussion on the existing methods implementing them.
In particular, we focus on the many choice points and subtleties a software engineer may encounter while developing a ML workflow.
Finally, we show a programming ecosystem namely, SciKit-learn plus some related Python libraries aimed at at supporting software engineers in implementing their workflow.