 This paper presents a comprehensive study on the use of machine learning for intrusion detection systems in the Internet of Things, IoT. It evaluates various feature extraction algorithms and multiple machine learning algorithms, including random forests, k-nearest neighbors, support vector machines, and stacked models. The authors also evaluate the performance of these algorithms when used together with the VGG16 model and DenseNet models. The results show that the combination of VGG16 and stacking achieve the best accuracy of 98.3%, demonstrating the potential of machine learning for intrusion detection in IoT environments. This article was authored by Daya Mazla, Mira Alatebi, Fod Alhidari, and others.