About Premise Reduction of Fuzzy Inference AlgorithmZhang, Cheng-yi; Niu, Qi; Peng, De-jun; Li, Juan
2009 Fuzzy Information and Engineering
doi: 10.1007/s12543-009-0019-8
AbstractThe comprehensive model with “weighted-objective nearness degree” is introduced in the process of multi-objective decision-making, by which a reduction problem of inference antecedents is studied in traditional fuzzy inference method. Moreover, based on the comprehensive model with “weighted-objective nearness degree”, SMTT fuzzy inference algorithm is proposed. This algorithm not only shows the relative importance of every antecedent component in fuzzy inference, but also considers the influence of nearness degree between every antecedent component's evaluation and inference objective on inference conclusions. The enactment of inference objective reflects the preference degree of decision maker to every antecedent component's evaluation. Therefore, it is much fitter for the demands of practical inference.
Linear Feature-weighted Support Vector MachineXing, Hong-jie; Ha, Ming-hu; Hu, Bao-gang; Tian, Da-zeng
2009 Fuzzy Information and Engineering
doi: 10.1007/s12543-009-0022-0
AbstractThe existing support vector machines (SVMs) are all assumed that all the features of training samples have equal contributions to construct the optimal separating hyperplane. However, for a certain real-world data set, some features of it may possess more relevances to the classification information, while others may have less relevances. In this paper, the linear feature-weighted support vector machine (LFWSVM) is proposed to deal with the problem. Two phases are employed to construct the proposed model. First, the mutual information (MI) based approach is used to assign appropriate weights for each feature of the whole given data set. Second, the proposed model is trained by the samples with their features weighted by the obtained feature weight vector. Meanwhile, the feature weights are embedded in the quadratic programming through detailed theoretical deduction to obtain the dual solution to the original optimization problem. Although the calculation of feature weights may add an extra computational cost, the proposed model generally exhibits better generalization performance over the traditional support vector machine (SVM) with linear kernel function. Experimental results upon one synthetic data set and several benchmark data sets confirm the benefits in using the proposed method. Moreover, it is also shown in experiments that the proposed MI based approach to determining feature weights is superior to the other two mostly used methods.
The Research of Ecological Security Evaluation for Mineral-resource Enterprises-a Case of ChinaZheng, Yun-hong; Li, Kai
2009 Fuzzy Information and Engineering
doi: 10.1007/s12543-009-0025-x
AbstractAs the rapid development of economy and society in the world, the situation of ecological security has become more and more severe. The problem on ecological ecurity is concerned with whether the continuous development of economy and society can realize or not, as well as the regional stability and international safety, so ecological security has become the primary measurement of the continuous development. Meanwhile there are a lot of research on ecological security of hometown and abroad, and the research range is becoming wider and in depth. With this background, this paper analyses the meaning of mineral-resource ecological security, sets up ecological security index system of mineral-resource enterprises based on the pressure-state-response (P-S-R) model, then determined the weights of the indices by analytical hierarchy process (AHP), and presents a method to evaluating it. After empirical analysis, the paper pointed out that evaluating the ecological environment security of mineral- resource enterprise is not only necessary but also the base of building ecological security alarming mechanism.