Stability of Time Series Models based on Fractional Order Weakening Buffer Operators

Different weakening butter operators in time series model analysis usually result in different model sensitivity, which sometimes affects the effectiveness of relevant operator-based methods. In this paper, the stability of two classic weakening buffer operator-based series models is studied; then a new data preprocessing method based on a novel fractional bidirectional weakening buffer operator is provided, whose effect in improving model stability is tested and utilized in prediction problems. Practical examples are employed to demonstrate the efficiency of the proposed method in improving model stability in noise scenarios. The comparison indicates that the proposed method overcomes the disadvantage of many weakening buffer operators in too subjectively biased weighting the new or the old information in forecasting. These expand the application of the proposed method in time series analysis