[II-78] Applicability of artificial neural network to prediction of waves caused by typhoon
Keywords:wave prediction, typhoon, Artificial neural network, machine learning, frequency spectrum
The applicability of neural networks to the prediction of waves caused by typhoons was discussed. In this study, it is confirmed whether observation data alone is sufficient for constructing the network, and whether waves caused by typhoons on routes that have not been experienced before can be predicted. The learning data of the waves by a typhoon of a certain route and the test data of waves by typhoon of a route slightly deviated from learning data were prepared, it was examined the prediction accuracy by neural network. As a result, a slight deviation of the typhoon route decreased prediction accuracy, so that the usefulness of using analysis data of various routes of typhoon was suggested.
Abstract password authentication.
Password is required to view the abstract. Please enter a password to authenticate.