Time Series Prediction as a Problem of Missing Values: Application to ESTSP2007 and NN3 Competition Benchmarks

In this paper, time series prediction is considered as a problem of missing values. A method for the determination of the missing time series values is presented. The method is based on two projection methods: a nonlinear one (Self-Organized Maps) and a linear one (Empirical Orthogonal Functions). The presented global methodology combines the advantages of both methods to get accurate candidates for the prediction values. The methods are applied to two time series competition datasets.