Decision-Making in Machine Learning

The flood of machine learning and deep learning model approaches increases the difficulty for companies and individuals to choose the best alternative for the respective project. The paper focuss on the problem of decision-making in choosing the best machine learning or deep learning model approach for the respective problem to be solved. At first the research gaps as well as problematic courses of action that lead to an irrational decision are identified. Then the data science quality model that provides an extensive overview of criteria for measuring and evaluating data science projects, lifecycles phases, and other components is proposed. Furthermore, the concept of decision theory and the designed model artifact are applied to the identified problem to demonstrate how a rational decision process in the field of data science should look like.