Neural Networks And Sea Time Series

This self-contained book is devoted to the application of neural networks to the concrete problem of handling the time series of sea data, i.e., significant wave heights and sea levels. Beginning with an explanation of neural networks fundamentals for beginners, the authors follow with a review of general statements and rigorous facts, and conclude with applications and algorithms for particular data sets having well-defined parameters.Specifically analyzed are the time series of measures observed from a system of sea buoys, as well as the reconstruction of missing values that are largely present in these time series. Neural networks are used to solve this problem, and the reconstructed time series are modeled to estimate the probability of large events using extreme value theory. This ability to model and estimate is important for the construction of sea structures, ports, and marine experiments."Neural Networks and Sea Time Series" is a unique monograph on the topic and may stimulate new research and results in the field. The book will be a useful resource for a diverse audience of graduate students, researchers and practitioners in applied mathematics, data analysis, meteorology, hydraulic, civil and marine engineering.