When Smart Cities Get Smarter via Machine Learning: An In-Depth Literature Review

• Published in: IEEE access Vol. 10; pp. 60985 - 61015
• Main Authors: Band, Shahab S., Ardabili, Sina, Sookhak, Mehdi, Chronopoulos, Anthony Theodore, Elnaffar, Said, Moslehpour, Massoud, Csaba, Mako, Torok, Bernat, Pai, Hao-Ting, Mosavi, Amir
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Accuracy; Algorithms; artificial intelligence; Artificial neural networks; big data; Data science; Decision trees; Deep learning; ensemble; Literature reviews; Machine learning; Medical services; Neural networks; Performance evaluation; Regression tree analysis; Smart cities; Smart city; smart grid; State-of-the-art reviews; Support vector machines; Taxonomy; Urban areas
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2022.3181718

The manuscript represents a comeprehensive and systematic literature review on the machine learning methods in the emerging applications of the smart cities. Application domains include the essential aspects of the smart cities including the energy, healthcare, transportation, security, and pollution. The research methodology presents a state-of-the-art taxonomy, evaluation and model performance where the ML algorithms are classified into one of the following four categories: decision trees, support vector machines, artificial neural networks, and advanced machine learning methods, i.e., hybrid methods, ensembles, and Deep Learning. The study found that the hybrid models and ensembles have better performance since they exhibit both a high accuracy and low overall cost. On the other hand, the deep learning (DL) techniques had a higher accuracy than the hybrid models and ensembles, but they demanded relatively higher computation power. Moreover, all these advanced ML methods had a slower processing speed than the single methods. Likewise, the support vector machine (SVM) and decision tree (DT) generally outperformed the artificial neural network (ANN) for accuracy and other metrics. However, since the difference was negligible, it can be concluded that using either of them is appropriate.