Guarding the Future: Anomaly Detection in IoT-Enabled Smart Cities

The Internet of Things (IoT) and its applications have developed into smarter, more linked systems that are used in every element of smart cities. The machine learning (ML) method is helpful to the further improvement in the varied intelligence and potential of the application as the amount of composed data increases. Researchers have been interested in smart transportation applications, which have been approached using both ML and IoT methodologies. This research suggested an innovative, effective detection system that depends on machine learning techniques to identify IoT attacks and stop harmful activity. Additionally, the UNSW-NB15 and CICIDS2017 datasets are utilized in this work. Initially, the data is pre-processed using combinations of several preprocessing and normalization processes. Then the features are extracted from the pre-processed data using an improved principal component analysis algorithm and genetic algorithm, and the attacks are classified using a random forest classifier. This study evaluates a variety of ML techniques for binary classification issues. The results show that for both datasets in the suggested model, the proposed random field technique is superior to conventional ML algorithms for attack detection.