Synthesizing judgment matrix and risk-odds matrix for small-sample combined forecasting

There are some small-sample cases in economic forecast, for lacking of quantitative information, in which it happens that some given forecasting method fits well while behaves badly in forecasting, while to the contrary, the situation behaves well when we make the combination of both qualitative and quantitative methods in forecasting. The concept of risk of forecasting is defined, as well as the concept and estimating method of risk-adds matrix. Then, taking the judgment matrix as the basic matrix and the risk-odds matrix as the adjustment one, the modeling strategy of forecasts combination which is obtained by dynamic synthesizing judgment matrix and risk-odds matrix is presented. In this strategy, as the mumble of samples changing, the advantages of both AHP and odds-matrix method can compensate each other, and the performance of small-sample combined forecasting can be improved.