Comparative Analysis of Anomaly Classification Method for Flood Prediction in Smart City Platform

The term “smart city” refers to a city that addresses issues using information and communication technology. Through the utilization of this information technology, smart cities can become data producers as well as data consumers. All the data produced by the smart city is gathered onto a single data platform. The information that was gathered can be applied to the study of a variety of urban issues, including floods, which is one of them. The sensors used to forecast the weather can also be utilized to look for unusual patterns, which can help with flood prediction. This work presents a comparative analysis to evaluate which anomaly classification algorithm is the most effective in predicting flooding. For this investigation, we use data obtained from an actual weather radar installed in a city in Indonesia. In addition, we compared our findings to those of any other readily available datasets. Based on the results, we can choose several classification algorithm with the best selected evaluation metric.