Today, the Internet of Things (IoT) is experiencing exponential growth in data generation and diverse network architectures. Numerous interconnected devices transmit data from sensor networks to large-scale networks. This data is vulnerable to various factors that can impact its efficiency, the functionality of the network, and thus the required Quality of Service (QoS). However, due to adverse conditions or malfunctioning equipment, the data collected will be anomalous. In this context, the anomaly detection problem is one of the key challenges requiring further research and tailored solutions. What is more anomaly detection is a critical security feature that identifies situations where the system behavior deviates from the expected norm, enabling the anomaly to be identified and remediated immediately. Implementing an efficient anomaly detection technique is essential to ensure service quality. Significant advancements in recent years have introduced new techniques based on machine learning and deep learning. These techniques are capable of identifying various threats in IoT systems, including zero-day attacks. To address these challenges, this study reviews the literature on anomaly detection in IoT networks through the lens of machine learning and deep learning. This chapter provides an overview of IoT networks, including their components, the types of data they generate, and the various anomaly types found within these networks. Additionally, we present and classify the latest anomaly detection techniques, including machine learning, deep learning, and hybrid approaches. A comparative analysis and a discussion of the datasets commonly used in the IoT domain are conducted. Finally, we examine the challenges and limitations faced in anomaly detection within IoT networks.