Abstract: Aiming at difficulties in selecting customer collaborative innovation partner,a partner selection model based on Support Vector Machine and Rough Sets theory has been proposed in this paper to solve the problem that decision-making attributes were varies and the sample data used for decision-making and analysis were insufficient.The key point of this model is the use of rough sets theory to pick out the important attributes as the pretreatment of data so that unimportant attributes in decision table can be deleted.Then support vector machine is used for customer classfication because of its advantage of doing better than others on dealing with small sample sets and non-liner problems.This method which not affect classification performance can reduce the data dimensions and the complexity in classification process.Finally,this method is applied to a real engineering case.The result of application proves the feasibility and availability of this method.