"Machine Learning (ML) methods are increasingly being used as alternatives or surrogates to scientific models to explain real-world phenomena in a number of disciplines. However, given the limited ability of "black-box" ML methods to learn generalizable and scientifically consistent patterns from limited volumes of data, there is a growing realization in the scientific and data science communities to incorporate scientific knowledge in the ML process. This emerging paradigm combining scientific knowledgeand data at an equal footing is labeled Science-Guided ML (SGML). By using scientific consistency as an essential criterion for assessing generalizability of ML models, SGML aims to go far and beyond conventional standards of black-box ML in modeling scientific systems. SGML also aims to accelerate scientific discovery using data by informing scientific models with better estimates of latent quantities, augmenting modeling components, and/or discovering new scientific laws. Knowledge Guided Machine Learning: Accelerating Discovery using Scientific Knowledge and Data provides an introduction to this rapidly growing field by discussing some of the common themes of research in SGML, using illustrative examples and case studies from diverse application domains and research communities as contributed book chapters"--
