Multiscale Polynomial Autoregressive Model for Monthly Sardines Catches Forecasting

The aim of this paper is to find a model to forecast 1-month ahead monthly sardines catches using a multivariate polynomial model combined with multi-scale stationary wavelet decomposition. The observed monthly sardines catches are decomposed into various sub-series employing wavelet decomposition and then appropriate sub-series are used as inputs to the autoregressive forecasting model. The forecasting strategy parameters are estimated using the least squares method and we find that the proposed forecaster achieves 99% of the explained variance with a MAPE below 7.6%.