This chapter introduces the various families of machine learning (ML) methods to analyze data collected from wearable devices. The chapter provides to the reader an outline of the typical data streams captured by wearable sensors. These data streams can provide to the user, research community and the wearables industry, useful insights about health behavior and health-related measures, forming the foundation for subsequent analysis. The chapter then examines, both supervised and unsupervised ML techniques, highlighting which methods are most suitable for different analysis goals. As far as supervised methods are concerned, classification and regression methods are examined separately with practical use-case examples to illustrate how these techniques can be applied to wearable data. Additionally, the chapter explains two unsupervised methods: association rule mining and clustering, which are used to uncover hidden patterns and group similar behaviors. The importance of deep learning and reinforcement learning is discussed later in this chapter. Location based and temporal algorithms are covered because they analyze location-specific and time-dependent patterns, which are important in wearables analysis. The chapter concludes by discussing considerations for enabling contextual relevance and interpretability of algorithms, so that the insights deriving from wearable data are actionable, and meaningful in real-world applications.