The field of polymer informatics leverages artificial intelligence (AI) and machine learning (ML) tools to design polymers with desired properties. This book chapter reviews advancements in developing efficient AI/ML models that can accurately predict the properties of a wide variety of polymers, thus accelerating the polymer discovery process by screening candidates that meet the target properties. It explores the origins and details of different polymer representations or fingerprints, including chemical-intuition-based, graph-based, and transformer-based, that serve as inputs to such ML models. The chapter emphasizes the importance of incorporating polymer invariances and constraints into these fingerprints to enhance the accuracy and efficiency of the ML models. Success stories illustrating the contribution of polymer informatics to the discovery of high-performing polymers are highlighted. Finally, the chapter provides an overall outlook on this field and discusses some of the remaining open challenges.