The self-tuning prediction of nonstationary time series and its application. (Chinese. English summary) Zbl 0627.62093
The self-tuning prediction of a class of nonstationary time series is considered in this paper. The solution explored is to add a constant disturbance term to a CARMA model, and on-line identification of the model is used to implement the adaptive prediction. An example of short term load forecasting is presented which shows the effectiveness of the method.
MSC:
62M20 | Inference from stochastic processes and prediction |