TY - JOUR AU - Richetti, Jonathan AU - Giron Cima, Elizabeth AU - Johann, Jerry Adriani AU - Uribe-Opazo, Miguel Angel PY - 2018/03/05 Y2 - 2024/03/29 TI - Data Mining Techniques for Rainfall Regionalization in Parana State JF - Acta Iguazu JA - Rev. Act. Igu. VL - 7 IS - 1 SE - ARTIGOS CIENTÍFICOS DO - 10.48075/actaiguaz.v7i1.17777 UR - https://e-revista.unioeste.br/index.php/actaiguazu/article/view/17777 SP - 1-8 AB - <p>The prevalence of agro-meteorological data for specific regions serve as parameters for agricultural and related climate studies. This study aims to regionalize the rainfall in the State of Paraná (Southern Brazil) through data mining techniques with ECMWF (European Centre for Medium Range Weather Forecasts) data from 1989 to 2013. The algorithms k-means and Simple EM (Expectation Maximization) for clustering were applied in Weka software, version 3.6. The quality of the clustering was determined with the J48 classification algorithm applied using training set. The decision tree presents similarity indexes and errors measures to determine the best number of cluster for this case. As results 6 regions of homogeneous rainfall in the state of Paraná were presented.</p> ER -