The combination of Machine Learning (ML) and the Internet of Things (IoT) is bringing major changes to many sectors by enabling smart automation and decision-making based on real-time data. ML techniques help IoT systems to handle large volumes of data efficiently, which leads to better performance and lower resource use. For instance, in smart factories, ML with IoT improves productivity and reduces waste. However, challenges like data privacy and security still need to be addressed for smooth adoption. This chapter gives a detailed overview of how different ML methods—such as supervised, unsupervised, reinforcement, deep, and federated learning—are used in IoT environments. Each method is explained with its applications: supervised learning is used for predicting maintenance needs and health diagnostics; unsupervised learning helps in detecting unusual behaviour; reinforcement learning works well in changing environments like traffic systems; deep learning handles complex data such as images and voice; and federated learning supports privacy by training models on devices without sharing raw data. The chapter also includes case studies from smart cities and healthcare, showing how ML improves safety and efficiency. It highlights key issues like interoperability and scattered data, and discusses solutions like differential privacy and blockchain to protect data and user trust. Finally, the chapter looks at future directions, including lightweight ML models, explainable AI (XAI), and flexible, privacy-focused architectures that will help develop the next generation of secure and intelligent IoT systems.