Wireless Sensor Networks (WSNs) are vital for various crucial applications, including environmental monitoring and industrial automation. Nevertheless, these types of networks usually discover anomalies the old-fashioned way which is by using static rules after the occurrence of events, hence making it slower to identify and respond to threats that may compromise the security and integrity of a network. The research suggests a machine learning system employed for real-time anomaly detection in WSNs. This system uses smart machine learning algorithms that constantly monitor network traffic data, thereby allowing swift abnormality identification. The proposed method increases the response and precision of detecting anomalies to enhance the overall security and reliability of WSNs. By experimenting with it, it has proved how this technique is better than conventional rule-based systems with improved accuracy and ability to identify many different kinds of anomalies without making many wrong decisions.