Wavelet Denoising Based Multivariate Polynomial For Anchovy Catches Forecasting

Abstract

In this paprer, a multivariate polynomial (MP) combined with denoising techniques is proposed to forecast 1-month ahead monthly anchovy catches in the north area of Chile. The anchovy catches data is denoised by using discrete stationary wavelet transform and then appropriate is used as inputs to the MP. The MP's parameters are estimated using the penalized least square (LS) method and the performance evaluation of the proposed forecaster showed that a 98% of the explained variance was captured with a reduced parsimony.