The swift advancement of the Internet of Things (IoT) has ushered in a wealth of benefits, allowing countless interconnected devices to interact and exchange data effortlessly. Previously, network traffic including unusual patterns, was mainly produced by established, secure endpoints with strong security features, like smartphones. With the advent of the Internet of Things (IoT) no matter how small or intricate device, now has the capability to produce unusual levels of network activity. One of the biggest challenges facing the IoT industry is network traffic, which can have a negative impact on the overall performance of IoT devices and systems. To address this issue, a random forest classifier has been developed specifically for classifying IoT data. Extra Trees offer a significant benefit by minimizing bias. This is achieved by randomly sampling from the entire dataset when building the trees. Random Forest is a widely recognized machine learning technique which favours accuracy, reliability, flexibility and scalability. The process of data preprocessing involves transforming unrefined data into a refined dataset. Chi-square based feature extraction is utilized to extract relevant information and this technique enhances classification by selecting the most important features from the extraction regions. In the end, the chosen characteristics are inputted into both an extra tree and random forest classifier to ensure precise categorization and the implementation of this endeavor is carried out utilizing Python programming.