Classification, Automation, and New Media: Proceedings of the 24th Annual Conference of the Gesellschaft f?r Klassifikation e.V., University of Passau [Paperback]

Given the huge amount of information in the internet and in practically every domain of knowledge that we are facing today, knowledge discovery calls for automation. The book deals with methods from classification and data analysis that respond effectively to this rapidly growing challenge. The interested reader will find new methodological insights as well as applications in economics, management science, finance, and marketing, and in pattern recognition, biology, health, and archaeology.1: Data Analysis, Statistics, and Classification.- Clustering and Models.- Some New and Some Old Results for the Polytomous Rasch Model.- Some Basic Results on the Extension of Quasi-likelihood Based Measurement Error Correction to Multivariate and Flexible Structural Models.- On the Optimal Number of Clusters in Histogram Clustering.- The Dual Dynamic Factor Analysis Models.- Joint Non-symmetrical Correspondence Analysis with Ordered Categories.- Regression Analysis of Extremely Multicollinear Data.- Modern Data Analysis: A Clash of Paradigms.- Two-mode Clustering with Genetic Algorithms.- Exact Tests for the Comparison of Binary Data Structures in Time.- Connected Maximum Split Clustering of Ladder Graphs.- A Generalization of Two-mode Three-way Asymmetric Multidimensional Scaling.- Central Limit Theorem for Probabilities of Correct Classification.- A New Model for the Analysis of Multitrait-multimethod Data.- Discrete and Continuous Models for Two-way Data.- 2: Pattern Recognition and Automation.- Towards Feature Fusion  A Classifier on the Basis of Automatically Generated Significant Contour Sections.- Convex Discriminant Analysis Tools for Non Convex Pattern Recognition.- Sparse Kernel Feature Analysis.- 3: Data Mining, Information Processing, and Automation.- Exploring Association Rules by Interactive Graphics.- Classification of Texts with Support Vector Machines: An Examination of the Efficiency of Kernels and Data-transformations.- Text Mining with the Help of Cohesion Tl£$