This book provides an accessible introduction and practical guidelines to apply asymmetric multidimensional scaling, cluster analysis, and related methods to asymmetric one-mode two-way and three-way asymmetric data. A major objective of this book is to present to applied researchers a set of methods and algorithms for graphical representation and clustering of asymmetric relationships. Data frequently concern measurements of asymmetric relationships between pairs of objects from a given set (e.g., subjects, variables, attributes,...), collected in one or more matrices. Examples abound in many different fields such as psychology, sociology, marketing research, and linguistics and more recently several applications have appeared in technological areas including cybernetics, air traffic control, robotics, and network analysis. The capabilities of the presented algorithms are illustrated by carefully chosen examples and supported by extensive data analyses. A review of the specialized statistical software available for the applications is also provided. This monograph is highly recommended to readers who need a complete and up-to-date reference on methods for asymmetric proximity data analysis.

For the first time in one text, this handy pedagogical reference presents comprehensive inference strategies for organizing disparate nonparametric statistics topics under one scheme, illustrating ways of analyzing data sets based on generic notions of proximity (of "closeness") between objects. Assignment Methods in Combinatorial Data Analysis specifically reviews both linear and quadratic assignment models ... covers extensions to multiple object sets and higher-order assignment indices ... considers methods of applying linear assignment models in common data analysis contexts ... discusses a second motion of assignment (or "matching") based upon pairs of objects ... explores confirmatory methods of augmenting multidimensional sealing, cluster analysis, and related techniques ... labels sections in order of priority for continuity and convenience ... and includes extensive bibliographies of related literature. Assignment Methods in Combinatorial Data Analysis gives authoritative coverage of statistical testing, and measures of association in a single source. It is required reading and an invaluable reference for researchers and graduate students in the behavioral and social sciences using quantitative methods of data representation. Book jacket.

The NATO Advanced Study Institute on Numerical Taxonomy took place on the 4th - 16th of July, 1982, at the Kur- und Kongresshotel Residenz in Bad Windsheim, Federal Republic of Germany. This volume is the proceedings of that meeting, and contains papers by over two-thirds of the participants in the Institute. Numerical taxonomy has been attracting increased attention from systematists and evolutionary biologists. It is an area which has been marked by debate and conflict, sometimes bitter. Happily, this meeting took place in an atmosphere of "GemUtlichkeit", though scarcely of unanimity. I believe that these papers will show that there is an increased understanding by each taxonomic school of each others' positions. This augurs a period in which the debates become more concrete and specific. Let us hope that they take place in a scientific atmosphere which has occasionally been lacking in the past. Since the order of presentation of papers in the meeting was affected by time constraints, I have taken the liberty of rearranging them into a more coherent subject ordering. The first group of papers, taken from the opening and closing days of the meeting, debate philosophies of classification. The next two sections have papers on congruence, clustering and ordination. A notable concern of these participants is the comparison and testing of classifications. This has been missing from many previous discussions of numerical classification.