The concept of the degree of similarity between interval-valued intuitionistic fuzzy sets （IVIFSs） is introduced, and some distance measures between IVIFSs are defined based on the Hamming distance, the normalized H...The concept of the degree of similarity between interval-valued intuitionistic fuzzy sets （IVIFSs） is introduced, and some distance measures between IVIFSs are defined based on the Hamming distance, the normalized Hamming distance, the weighted Hamming distance, the Euclidean distance, the normalized Euclidean distance, and the weighted Euclidean distance, etc. Then, by combining the Hausdorff metric with the Hamming distance, the Euclidean distance and their weighted versions, two other similarity measures between IVIFSs, i. e., the weighted Hamming distance based on the Hausdorff metric and the weighted Euclidean distance based on the Hausdorff metric, are defined, and then some of their properties are studied. Finally, based on these distance measures, some similarity measures between IVIFSs are defined, and the similarity measures are applied to pattern recognitions with interval-valued intuitionistic fuzzy information.展开更多
为了进一步改善和提高基于模式的时间序列趋势相似性度量效果,在时间序列分段线性表示的基础上,依据分段子序列的均值及其线性拟合函数的导数符号,实现时间序列的分段模式化,以模式之间的异同性定义模式匹配距离,借鉴动态时间弯曲(Dynam...为了进一步改善和提高基于模式的时间序列趋势相似性度量效果,在时间序列分段线性表示的基础上,依据分段子序列的均值及其线性拟合函数的导数符号,实现时间序列的分段模式化,以模式之间的异同性定义模式匹配距离,借鉴动态时间弯曲(Dynamic Time Warping,DTW)的动态规划原理,提出一种动态模式匹配方法(Dynamic Pattern Matching,DPM)。实验结果表明,该方法能够在不同压缩率条件下,准确度量等长时间序列的趋势相似性,而且时间消耗较低。时间序列不等长作为存在数据缺失的一种表现形式,该方法的度量效果与数据缺失比例之间的关系值得进一步的深入研究。展开更多
基金The National Natural Science Foundation of China (No70571087)the National Science Fund for Distinguished Young Scholarsof China (No70625005)
文摘The concept of the degree of similarity between interval-valued intuitionistic fuzzy sets （IVIFSs） is introduced, and some distance measures between IVIFSs are defined based on the Hamming distance, the normalized Hamming distance, the weighted Hamming distance, the Euclidean distance, the normalized Euclidean distance, and the weighted Euclidean distance, etc. Then, by combining the Hausdorff metric with the Hamming distance, the Euclidean distance and their weighted versions, two other similarity measures between IVIFSs, i. e., the weighted Hamming distance based on the Hausdorff metric and the weighted Euclidean distance based on the Hausdorff metric, are defined, and then some of their properties are studied. Finally, based on these distance measures, some similarity measures between IVIFSs are defined, and the similarity measures are applied to pattern recognitions with interval-valued intuitionistic fuzzy information.
基金Supported by the Natural Science Foundation of Fujian Province of China under Grant No.A0510020(福建省自然科学基金)the Int'I Science and Technology Cooperation Project of Fujian Province of China under Grant No.20041014(福建省国际科技合作项目)
文摘为了进一步改善和提高基于模式的时间序列趋势相似性度量效果,在时间序列分段线性表示的基础上,依据分段子序列的均值及其线性拟合函数的导数符号,实现时间序列的分段模式化,以模式之间的异同性定义模式匹配距离,借鉴动态时间弯曲(Dynamic Time Warping,DTW)的动态规划原理,提出一种动态模式匹配方法(Dynamic Pattern Matching,DPM)。实验结果表明,该方法能够在不同压缩率条件下,准确度量等长时间序列的趋势相似性,而且时间消耗较低。时间序列不等长作为存在数据缺失的一种表现形式,该方法的度量效果与数据缺失比例之间的关系值得进一步的深入研究。
