By Reinhold Decker, Hans-Joachim Lenz

ISBN-10: 3540709800

ISBN-13: 9783540709800

The ebook makes a speciality of exploratory information research, studying of latent constructions in datasets, and unscrambling of information. It covers a wide diversity of equipment from multivariate statistics, clustering and type, visualization and scaling in addition to from info and time sequence research. It offers new ways for info retrieval and information mining. moreover, the booklet stories not easy purposes in advertising and administration technological know-how, banking and finance, bio- and healthiness sciences, linguistics and textual content research, statistical musicology and sound class, in addition to archaeology. unique emphasis is wear interdisciplinary learn and the interplay among thought and perform.

Show description

Read or Download Advances in data analysis: proceedings of the 30th Annual Conference of The Gesellschaft fur Klassifikation e.V., Freie Universitat Berlin, March 8-10, 2006 PDF

Best organization and data processing books

Get The Statistical Sleuth: A Course in Methods of Data Analysis PDF

While you're into facts an outstanding publication but when you're taking it since you need to now not any such strong learn.

Computing and Information Technologies by George Antoniou, Dorothy Deremer PDF

Provides a balanced presentation of the newest thoughts, algorithms and functions in machine technology and engineering. The papers written, supply a automobile for brand new examine and improvement.

Additional info for Advances in data analysis: proceedings of the 30th Annual Conference of The Gesellschaft fur Klassifikation e.V., Freie Universitat Berlin, March 8-10, 2006

Sample text

If the clustering algorithm is based on a distance matrix then, in validation, indexes based on the inertia and indexes based on a distance matrix are allowed. If an algorithm designed strictly for symbolic data is used then for validation indexes based on inertia and “symbolic” indexes are most appropriate. Thus, four paths of classification procedure may be distinguished: 34 Andrzej Dudek Fig. 1. Clustering method and cluster quality indexes for symbolic data. (Source: Own research based on Verde (2004), Chavent at al.

2006): Model Selection for the Binary Latent Class Model. A Monte Carlo Simulation. In: V. -H. Bock, A. Ferligoj and A. ): Data Science and Classification. Springer, Berlin, 91–99. G. and WILLEKENS, F. (2005): Model-based Clustering of Sequential Data with an Application to Contraceptive Use Dynamics. Mathematical Population Studies, 12, 135–157. R. B. (2001): Testing the Number of Components in a Normal Mixture. Biometrika, 88, 767-778. J. and PEEL, D. (2000): Finite Mixture Models. John Wiley & Sons, New York.

Indexes designed for symbolic data: symbolic inertia and homogeneity based quality index can also be used for symbolic cluster validation but the results may be worse than those achieved by using the Hubert and Levine or the Baker and Hubert index. 6 Final remarks In this paper several cluster quality indexes were compared on 100 artificially generated symbolic data sets. The experiment showed that the most adequate ones for this kind of data are the Hubert and Levine and the Baker and Hubert indexes.

Download PDF sample

Advances in data analysis: proceedings of the 30th Annual Conference of The Gesellschaft fur Klassifikation e.V., Freie Universitat Berlin, March 8-10, 2006 by Reinhold Decker, Hans-Joachim Lenz


by Ronald
4.2

New PDF release: Advances in data analysis: proceedings of the 30th Annual
Rated 4.72 of 5 – based on 44 votes