方法对比

	模糊 C均值聚类 (FCM)	粒计算（信息粒化）	谱聚类
领域	机器学习	软计算	机器学习
方法族	Machine learning	Machine learning	Machine learning
起源年份	1981	1997	2002
提出者	Joseph Dunn; James Bezdek	Lotfi A. Zadeh (information granulation); developed by Pedrycz, Skowron, Yao	Ng, A. Y.; Jordan, M. I.; Weiss, Y.
类型	Soft (fuzzy) partitional clustering	Framework for multi-granularity information processing	Graph-based clustering (spectral method)
开创性文献	Dunn, J. C. (1973). A fuzzy relative of the ISODATA process and its use in detecting compact well-separated clusters. Journal of Cybernetics, 3(3), 32–57.	Zadeh, L. A. (1997). Toward a theory of fuzzy information granulation and its centrality in human reasoning and fuzzy logic. Fuzzy Sets and Systems, 90(2), 111–127.	Ng, A. Y., Jordan, M. I., & Weiss, Y. (2002). On Spectral Clustering: Analysis and an Algorithm. Advances in Neural Information Processing Systems, 14, 849–856.
别名	FCM, fuzzy clustering, soft k-means, bulanık c-ortalama kümeleme	information granulation, computing with granules, three-way granular computing, tanecikli hesaplama	NJW spectral clustering, graph Laplacian clustering, normalized spectral clustering, spectral graph clustering
相关	3	3	5
摘要	Fuzzy C-Means is a soft clustering algorithm in which every data point belongs to every cluster with a graded membership between 0 and 1, rather than being assigned to exactly one cluster. Originated by Joseph Dunn in 1973 and generalized by James Bezdek in 1981, it minimizes a fuzzy-weighted within-cluster variance, making it well suited to data whose groups overlap or have no sharp boundaries.	Granular computing is a problem-solving paradigm that processes information in 'granules' — clumps of objects drawn together by indistinguishability, similarity, or functionality — rather than at the level of individual data points. Articulated by Lotfi Zadeh in 1997 as fuzzy information granulation and developed into a broad framework, it provides a unifying umbrella over fuzzy sets, rough sets, and interval methods, letting analysis move to whichever level of detail a problem actually requires.	Spectral Clustering is a graph-based unsupervised learning algorithm, formalized by Ng, Jordan, and Weiss in 2002, that maps data points into a low-dimensional eigenspace derived from the similarity graph's Laplacian before applying k-means. This spectral embedding makes it possible to recover clusters of arbitrary shape — rings, crescents, interleaved spirals — that Euclidean distance-based methods consistently fail to separate.