Forecasting stochastic processes using singular spectrum analysis: Aspects of the theory and application