Abstract: To solve the problem of low jamming decision accuracy of template matching (TM) method under condition of unknown jamming rule base, a method based on transfer component analysissupport vector machine (TCASVM) is proposed. For airborne multifunctional fire control radar on airtoair scene, the signal features are extracted and the jamming rule base and unknown threat dataset are constructed. The features of the two sample sets are mapped to the same low dimension hidden space via transfer component analysis. The hidden space sample features are extracted and trained by support vector machine to realize jamming decision on the unknown threat dataset. The experiment results show that the proposed method can effectively improve the jamming decision accuracy rate. The jamming decision problem under the condition of unknown jamming rule base is solved because of excellent learning and generalization ability of TCASVM.