Prediction of Tropical Storms Using Self-organizing Incremental Neural Networks and Error Evaluation

• Published in: Neural Information Processing Vol. 10636; pp. 846 - 855
• Main Authors: Kim, Wonjik, Hasegawa, Osamu
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2017; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Meteorology; Natural disaster; Self-organizing incremental neural network; Tropical storm
• ISBN: 9783319700892; 3319700898
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-70090-8_86

In this paper, we propose a route prediction method that uses a self-organizing incremental neural network (SOINN). For the training and testing of the neural network, only the latitude and longitude of the tropical storm and atmospheric information around East Asia are required. Our proposed method can predict the movement of a tropical storm with only a short calculation time, and the prediction accuracy is close to the accuracy of the Japan Meteorological Agency. This paper describes the algorithm used for the neural network training, the process for handling the data sets and the method used to predict the storm trajectory. Additionally, experimental results that indicate the performance of our method are presented in the results section.