Multiple Attribute Decision Analysis (MADA), known to be simple and convenient, is one of the most commonly used methods for Effectiveness Evaluation of Fighter (EEF), in which the attribute weight assignment plays a key role. Generally, there are two parts in the index system of MADA, i.e. performance index and decision index (or label), which denote the specific performance and the category of the object, respectively. In some index systems of EEF, the labels can be easily obtained, which are presented as the generations of fighters. However, the existing methods of attribute weight determination usually ignore or do not take full advantage of the supervisory function of labels. To make up for this deficiency, this paper develops an objective method based on fuzzy Bayes risk. In this method, a loss function model based on Gaussian kernel function is proposed to cope with the drawback that the loss function in Bayes risk is usually determined by experts. In order to evaluate the credibility of assigned weights, a longitudinal deviation and transverse residual correlation coefficient model is designed. Finally, a number of experiments, including the comparison experiments on University of California Irvine (UCI) data and EEF, are carried out to illustrate the superiority and applicability of the proposed method.