Aiming at the requirements of adaptive adjustment of jamming strategies according to radar behaviors and real-time evaluation during radar countermeasures， an evaluation method based on the fusion of AdaBoost algorithm and Naive Bayes classifier is proposed to meet the requirements of adaptive adjustment and real-time evaluation of interference strategy with radar behavior during radar countermeasures. This method extracts features from the time domain， frequency domain， modulation domain， and polarization domain of radar signals， using a naive Bayes classifier as the base classifier. Through AdaBoost iteration and dynamic feedback scaling adjustment of difficult to distinguish sample weights， the accuracy and stability of the model evaluation ability are enhanced. Simulation experiments have shown that the performance of this algorithm is superior to typical algorithms. This study provides an optimized solution for real-time quantitative evaluation of radar interference effects in complex electromagnetic environments， which can be applied to intelligent interference decision-making in radar countermeasures.