To improve the performance of multiple classifier system, a knowledge discovery based dynamic weighted voting （KD-DWV） is proposed based on knowledge discovery. In the method, all base classifiers may be allowed to ...To improve the performance of multiple classifier system, a knowledge discovery based dynamic weighted voting （KD-DWV） is proposed based on knowledge discovery. In the method, all base classifiers may be allowed to operate in different measurement/feature spaces to make the most of diverse classification information. The weights assigned to each output of a base classifier are estimated by the separability of training sample sets in relevant feature space. For this purpose, some decision tables （DTs） are established in terms of the diverse feature sets. And then the uncertainty measures of the separability are induced, in the form of mass functions in Dempster-Shafer theory （DST）, from each DTs based on generalized rough set model. From the mass functions, all the weights are calculated by a modified heuristic fusion function and assigned dynamically to each classifier varying with its output. The comparison experiment is performed on the hyperspectral remote sensing images. And the experimental results show that the performance of the classification can be improved by using the proposed method compared with the plurality voting （PV）.展开更多
Radar anti-jamming performance evaluation is a necessary link in the process of radar development,introduction and equipment. The applications of generalized rough set theory are proposed and discussed in this paper t...Radar anti-jamming performance evaluation is a necessary link in the process of radar development,introduction and equipment. The applications of generalized rough set theory are proposed and discussed in this paper to address the problems of big data, incomplete data and redundant data in the construction of evaluation index system. Firstly, a mass of real-valued data is converted to some interval-valued data to avoid an unacceptable number of equivalence classes and classification rules, and the interval similarity relation is employed to make classifications of this interval-valued data. Meanwhile, incomplete data can be solved by a new definition of the connection degree tolerance relation for both interval-valued data and single-valued data, which makes a better description of rough set than the traditional limited tolerance relation. Then, E-condition entropy-based heuristic algorithm is applied to making attribute reduction to optimize the evaluation index system, and final decision rules can be extracted for system evaluation. Finally, the feasibility and advantage of the proposed methods are testified by a real example of radar anti-jamming performance evaluation.展开更多
The model of grey multi-attribute group decision-making （MAGDM） is studied, in which the attribute values are grey numbers. Based on the generalized dominance-based rough set approach （G-DR- SA）, a synthetic secur...The model of grey multi-attribute group decision-making （MAGDM） is studied, in which the attribute values are grey numbers. Based on the generalized dominance-based rough set approach （G-DR- SA）, a synthetic security evaluation method is presented. With-the grey MAGDM security evaluation model as its foundation, the extension of technique for order performance by similarity to ideal solution （TOPSIS） integrates the evaluation of each decision-maker （DM） into a group＇s consensus and obtains the expected evaluation results of information system. Via the quality of sorting （QoS） of G-DRSA, the inherent information hidden in data is uncovered, and the security attribute weight and DMs＇ weight are rationally obtained. Taking the computer networks in a certain university as objects, the example illustrates that this method can effectively remove the bottleneck of the grey MAGDM model and has practical significance in the synthetic security evaluation.展开更多
To improve the performance of the multiple classifier system, a new method of feature-decision level fusion is proposed based on knowledge discovery. In the new method, the base classifiers operate on different featur...To improve the performance of the multiple classifier system, a new method of feature-decision level fusion is proposed based on knowledge discovery. In the new method, the base classifiers operate on different feature spaces and their types depend on different measures of between-class separability. The uncertainty measures corresponding to each output of each base classifier are induced from the established decision tables （DTs） in the form of mass function in the Dempster-Shafer theory （DST）. Furthermore, an effective fusion framework is built at the feature-decision level on the basis of a generalized rough set model and the DST. The experiment for the classification of hyperspectral remote sensing images shows that the performance of the classification can be improved by the proposed method compared with that of plurality voting （PV）.展开更多
基金This project was supported by the National Basic Research Programof China (2001CB309403)
文摘To improve the performance of multiple classifier system, a knowledge discovery based dynamic weighted voting （KD-DWV） is proposed based on knowledge discovery. In the method, all base classifiers may be allowed to operate in different measurement/feature spaces to make the most of diverse classification information. The weights assigned to each output of a base classifier are estimated by the separability of training sample sets in relevant feature space. For this purpose, some decision tables （DTs） are established in terms of the diverse feature sets. And then the uncertainty measures of the separability are induced, in the form of mass functions in Dempster-Shafer theory （DST）, from each DTs based on generalized rough set model. From the mass functions, all the weights are calculated by a modified heuristic fusion function and assigned dynamically to each classifier varying with its output. The comparison experiment is performed on the hyperspectral remote sensing images. And the experimental results show that the performance of the classification can be improved by using the proposed method compared with the plurality voting （PV）.
基金the Opening Project of the State Key Laboratory of Complex Electromagnetic Environment Effects on Electronics and Information System(No.CEMEE2014K0301A)
文摘Radar anti-jamming performance evaluation is a necessary link in the process of radar development,introduction and equipment. The applications of generalized rough set theory are proposed and discussed in this paper to address the problems of big data, incomplete data and redundant data in the construction of evaluation index system. Firstly, a mass of real-valued data is converted to some interval-valued data to avoid an unacceptable number of equivalence classes and classification rules, and the interval similarity relation is employed to make classifications of this interval-valued data. Meanwhile, incomplete data can be solved by a new definition of the connection degree tolerance relation for both interval-valued data and single-valued data, which makes a better description of rough set than the traditional limited tolerance relation. Then, E-condition entropy-based heuristic algorithm is applied to making attribute reduction to optimize the evaluation index system, and final decision rules can be extracted for system evaluation. Finally, the feasibility and advantage of the proposed methods are testified by a real example of radar anti-jamming performance evaluation.
文摘The model of grey multi-attribute group decision-making （MAGDM） is studied, in which the attribute values are grey numbers. Based on the generalized dominance-based rough set approach （G-DR- SA）, a synthetic security evaluation method is presented. With-the grey MAGDM security evaluation model as its foundation, the extension of technique for order performance by similarity to ideal solution （TOPSIS） integrates the evaluation of each decision-maker （DM） into a group＇s consensus and obtains the expected evaluation results of information system. Via the quality of sorting （QoS） of G-DRSA, the inherent information hidden in data is uncovered, and the security attribute weight and DMs＇ weight are rationally obtained. Taking the computer networks in a certain university as objects, the example illustrates that this method can effectively remove the bottleneck of the grey MAGDM model and has practical significance in the synthetic security evaluation.
文摘To improve the performance of the multiple classifier system, a new method of feature-decision level fusion is proposed based on knowledge discovery. In the new method, the base classifiers operate on different feature spaces and their types depend on different measures of between-class separability. The uncertainty measures corresponding to each output of each base classifier are induced from the established decision tables （DTs） in the form of mass function in the Dempster-Shafer theory （DST）. Furthermore, an effective fusion framework is built at the feature-decision level on the basis of a generalized rough set model and the DST. The experiment for the classification of hyperspectral remote sensing images shows that the performance of the classification can be improved by the proposed method compared with that of plurality voting （PV）.
