This book chapter explores the application of machine learning techniques in handball. Research in handball is a historically reliant on manual data annotation. However, recent advancements in ML, particularly in action recognition, performance prediction, tactical pattern recognition, expected goals modeling, and sequence detection, offer new possibilities for in-depth game analysis. By utilizing video, event, and sensor data, researchers and practitioners can now automate the extraction of complex game phases, predict future performances, and identify tactical patterns. The introduction of frameworks like FAUPA-ML demonstrates the potential to scale domain-specific knowledge to large datasets, enabling the analysis of large quantities of data without the limiting costs of manual annotation. This chapter underscores the importance of ML in enhancing our understanding of handball dynamics, providing tools for coaches and analysts to refine training processes and prepare strategically for opponents, and marking a shift towards data-driven approaches in sports analytics.