Wavelet additive forecasting model to support the fisheries industry

We present a forecasting strategy based on stationary wavelet decomposition combined with linear regression to improve the accuracy of one-month-ahead pelagic fish catches forecasting of the fisheries industry in southern zone of Chile. The general idea of the proposed forecasting model is to decompose the raw data set into long-term trend component and short-term fluctuation component by using wavelet decomposition. In wavelet domain, the components are predicted using a linear autoregressive model. Hence, proposed forecaster is the co-addition of two predicted components. We demonstrate the utility of the strategy on anchovy catches data set for monthly periods from 1978 to 2007. We find that the proposed forecasting scheme achieves a 98% of the explained variance with a reduced parsimonious.