This report describes an application of a hierarchical grouping procedure to a problem of grouping test profiles to maximize the homogeneity of profiles within the same clusters, taking account of all profile variables and all clusters at the same time. The data are 25 test profiles to which Sawrey, Keller, + Conger applied a different grouping technique. The results of the two grouping techniques are compared. Desirable characteristics of the hierarchical grouping technique are demonstrated.Any measure of profile similarity can be used. It is unnecessary to specify in advance the number of groups to be formed or to select nucleus groups. The cost of grouping can be evaluated in terms of any criterion expressed as a functional relation, or objective function. The resulting hierarchical structure of k profiles is that which, at each stage (k, k-1, ..., 1), least impairs the objective-function value. hus the hierarchical grouping technique shows not only the order in which profiles must be grouped so as to yield the optimal value of the objective function when the number of profiles is systematically reduced, but also the costs of the grouping at each stage of the hierarchy.

Includes Abstracts of technical reports issued by the Division and its antecedent organizations; issued separately as a supplement to some vols.

The study of the evolution of an urban system raises a number of practical questions, as to the operational definition of the components of the system as well as to the incorporation of the factor time in the analysis. In this chapter the fo cuss is on the urban centres, the nodes of the urban system. One of the first pro blems to be solved now is the definition of such a node. It has been argued above that population size can be used as an approximation of the complexity of the spatial economic structure of a region. The acceptation of this notion makes it possible to study the urban system as a population system. The analysis of the population system over the study period will be pursued from this perspective. The study period, 1840-1970, covers completely the industrialization and urbanization phase which occurred during the industrial revolution. Over this period the spatial organizational structure of society evolved from a rural-commercial to wards a modern urban-industrial state. One of the major events in this period of change is the spatial redistribution of the population and the population growth. This process will be examined below at three levels of aggregation: (I) the popu lation distribution as one frequency distribution, (2) the population distribution dis aggregated into a n~ber of size classes, and (3) the individual cities which make up the population distribution.

The two-volume set LNCS 4131 and LNCS 4132 constitutes the refereed proceedings of the 16th International Conference on Artificial Neural Networks, ICANN 2006. The set presents 208 revised full papers, carefully reviewed and selected from 475 submissions. This second volume contains 105 contributions related to neural networks, semantic web technologies and multimedia analysis, bridging the semantic gap in multimedia machine learning approaches, signal and time series processing, data analysis, and more.

A tremendous amount of work has been done over the last thirty years in cluster analysis, with a significant amount occurring since 1960. A substantial portion of this work has appeared in many journals, including numerous applied journals, and a unified ex position is lacking. The purpose of this monograph is to supply such an exposition by presenting a brief survey on cluster analysis. The main intent of the monograph is to give the reader a quick account of the prob lem of cluster analysis and to expose to him the various aspects thereof. With this intent in mind much detail has been omitted, particularly in so far as detailed examples are considered. Most of the references stated within the text contain examples and the reader can consult them for additional information on specific topics. Efforts were made to include in the reference section all papers that played a role in developing the "theory" of cluster analysis. Any omission of such references was not intentional and we would appreciate knowing about them. Many references to papers in applied journals are also contained, however, the list-is far from being complete. This monograph has been greatly influenced by the work of many people, most notably, J. A. Hartigan, D. Wishart, J. K. Bryan, R. E. Jensen, H. D. Vinod, and M. R. Rao. Several portions of the monograph were motivated by research performed under the support of NASA Manned Spacecraft Center, Earth Observations Division, under Contract NAS 9-12775.

The two volumes set LNCS 7653 and 7654 constitutes the refereed proceedings of the 4th International Conference on Computational Collective Intelligence, ICCCI, held in Ho Chi Minh City, Vietnam, in November 2012. The 113 revised full papers presented were carefully reviewed and selected from 397 submissions. The papers are organized in topical sections on (Part I) knowledge integration; data mining for collective processing; fuzzy, modal, and collective systems; nature inspired systems; language processing systems; social networks and semantic web; agent and multi-agent systems; classification and clustering methods; modeling and optimization techniques for business intelligence; (Part II) multi-dimensional data processing; web systems; intelligent decision making; methods for scheduling; collective intelligence in web systems – web systems analysis; advanced data mining techniques and applications; cooperative problem solving; computational swarm intelligence; and semantic methods for knowledge discovery and communication