To deal with the matter that airborne sensors may not acquire targets' parameters in combat time limitation, which leads to failure of attributes matching, a rough decision-making model for UCAV based on on-line correction time series forecasting method is presented. With the time series analysis results of sensors' data in previous time, the optimal autoregressive moving average model is constructed. With the acquired delayed data, sensors' data in future time, which would be used as an input to the process of attributes matching , are predicted and the data are corrected by on-line correction prediction method. As the analysis of an example of target threat assessment for UCAV shows, the decision-making model can extract rules in all conditions and present efficient decision-making advices.