Informed Machine Learning (Informed ML) refers to the idea of injecting additional prior knowledge into data-driven learning systems. Such knowledge can be given in various forms such as scientific equations or logic rules which provide relevant information about a problem domain or task at hand. Integrating prior knowledge at various stages of the machine learning pipeline can help to improve generalization and trustworthiness. Specifically, Informed ML can help to train models when training data is scarce or to ensure conformity with regulations or safety demands. In this introductory chapter, we briefly explain the concept of Informed ML, provide an overview of the chapters in this book, and categorize the contributed research and results with respect to a taxonomy of Informed ML.