Engenharia de features para construção de um modelo de previsão de eventos hidrometeorológicos em Curitiba

Abstract

Floods and urban flooding are becoming increasingly frequent, consequently, the number of people and infrastructure affected by these events has increased. For this reason, it is essential to have accurate models for predicting hydrometeorological events in order to reduce damages in advance. This thesis implemented a prediction model for this type of event in the city of Curitiba, using data from rain gauges, altimetry, and rivers and lakes, applying Feature Engineering techniques to understand which factors contribute most to the prediction of floods and urban flooding. The aim was to identify the variables that most contribute to predicting hydrometeorological events. It was observed that the variables that contributed most to a better model were rainfall data and altitude data. The classification model used in this work was able to assist in identifying the most important variables for predicting hydrometeorological events in Curitiba city. The knowledge generated can serve as a basis for more complex models to be implemented in future.