Open AccessUtilization of data mining classification techniques to identify the effect of Madden-Julian Oscillation on increasing sea wave height over East Java Waters
Author(s) -
Furqon Alfahmi,
Oky Sukma Hakim,
Rima Dewi,
A Khaerima
Publication year - 2019
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.179
H-Index - 26
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/399/1/012062
Subject(s) - java , madden–julian oscillation , naive bayes classifier , meteorology , wind speed , computer science , climatology , geology , machine learning , geography , support vector machine , convection , programming language
East Java BPBD data recorded 18 marine accidents in 2018, which increased by 1 event compared to the previous year. It is interesting to study the waters around East Java which are divided into 9 regions. The wind is a major factor in the high wave generation, but the contribution of weather phenomena triggered by the marine environment is important to identify. Phenomenon such as Madden-Julian Oscillation (MJO) has a cycle through the Indonesia territory, becomes a factor that should be suspected. MJO identification uses the Real-Time Multivariate MJO (RMM)-1 and RMM-2 index, which can be combined with the wind speed data using data mining classification techniques to get the thresholds value of wave height data obtained from the analysis of Windwave-05 model. The classification is helped by WEKA’s machine learning algorithm, by determining 4 selected classification algorithms including Naïve Bayes, J48, JRip, and Multi-Class Classifier. The data validation using the K-fold cross-validation method with a number of folds is 10 units. The accuracy value of the best algorithm obtained in each waters region ranges from 63.02% to 84.50%. The overall accuracy value increases by 0.24% to 4.41% compared to only using wind factors, except for the Waters of Bawean Island and Masalembu Islands.