This chapter examines the critical importance of a human-centered machine learning (HCML) approach in ensuring that the human remains at the core of the design and development process. Technical challenges in machine learning (ML) are strongly intertwined with ethical risks, requiring (a) methods to keep the human in the machine loop through participatory methods and well-designed interactions and (b) instruments for multidisciplinary teams to collaborate toward keeping the human at the center of every design and development decision. The chapter explores human-in-the-loop systems, interactive ML, and participatory ML methodologies, which prioritize user involvement to create adaptive systems responsive to real-world contexts. A framework for cross-functional collaboration is introduced, emphasizing the early and continuous inclusion of diverse stakeholders to ensure ethically sound and human-centered solutions. The framework is then illustrated in detail through a section dedicated to methods and tools presented with two case studies. The chapter closes by presenting the HCML heuristics as an actionable practice for the evaluation of ML from a human-centered perspective.