{"title":"Tidal Data Analysis using ANN","authors":"Ritu Vijay, Rekha Govil","country":null,"institution":"","volume":24,"journal":"International Journal of Information and Communication Engineering","pagesStart":4238,"pagesEnd":4242,"ISSN":"1307-6892","URL":"https:\/\/publications.waset.org\/pdf\/1909","abstract":"The design of a complete expansion that allows for\r\ncompact representation of certain relevant classes of signals is a\r\ncentral problem in signal processing applications. Achieving such a\r\nrepresentation means knowing the signal features for the purpose of\r\ndenoising, classification, interpolation and forecasting. Multilayer\r\nNeural Networks are relatively a new class of techniques that are\r\nmathematically proven to approximate any continuous function\r\narbitrarily well. Radial Basis Function Networks, which make use of\r\nGaussian activation function, are also shown to be a universal\r\napproximator. In this age of ever-increasing digitization in the\r\nstorage, processing, analysis and communication of information,\r\nthere are numerous examples of applications where one needs to\r\nconstruct a continuously defined function or numerical algorithm to\r\napproximate, represent and reconstruct the given discrete data of a\r\nsignal. Many a times one wishes to manipulate the data in a way that\r\nrequires information not included explicitly in the data, which is\r\ndone through interpolation and\/or extrapolation.\r\nTidal data are a very perfect example of time series and many\r\nstatistical techniques have been applied for tidal data analysis and\r\nrepresentation. ANN is recent addition to such techniques. In the\r\npresent paper we describe the time series representation capabilities\r\nof a special type of ANN- Radial Basis Function networks and\r\npresent the results of tidal data representation using RBF. Tidal data\r\nanalysis & representation is one of the important requirements in\r\nmarine science for forecasting.","references":null,"publisher":"World Academy of Science, Engineering and Technology","index":"Open Science Index 24, 2008"}