The Past and Present of Predictive Models for Anomaly Detection in Smart Cities: A Systematic Review

Compartir
Detecting anomalies in smart cities is a novel area that started being studied in the 21st century. This master’s thesis aims to find the most accurate predictive models that can be explainable to scholars and industry stakeholders. With that goal in mind, a PRISMA 2020 for systematic literature reviews methodology is approached to review the papers that have been published in Emerald Insights, IEEE Xplore, Science Direct, and Web of Science with the concepts of Smart Cities, Data Science, and Predictive Models between 2000 and the first half of 2023. The findings show that the algorithms that have been studied the most are for classification, supervised machine learning. This thesis not only took into account the theoretical part, but also attempted addressing those techniques by forecasting the energy consumption in buildings in Barcelona, classifying if those outcomes were an anomaly, and finally clustering to find the consumption patterns. The deliverables are disclosed in a ObservableHQ notebook and a dashboard in Google Data Studio ​
Aquest document està subjecte a una llicència Creative Commons:Reconeixement - No comercial - Sense obra derivada (by-nc-nd) Creative Commons by-nc-nd4.0

Localització