Showing papers on "Cluster analysis published in 1968"
IBM1
TL;DR: This paper reviews statistical, adaptive, and heuristic techniques used in laboratory investigations of pattern recognition problems and includes correlation methods, discriminant analysis, maximum likelihood decisions minimax techniques, perceptron-like algorithms, feature extraction, preprocessing, clustering and nonsupervised learning.
Abstract: This paper reviews statistical, adaptive, and heuristic techniques used in laboratory investigations of pattern recognition problems. The discussion includes correlation methods, discriminant analysis, maximum likelihood decisions minimax techniques, perceptron-like algorithms, feature extraction, preprocessing, clustering and nonsupervised learning. Two-dimensional distributions are used to illustrate the properties of the various procedures. Several experimental projects, representative of prospective applications, are also described.
317 citations
TL;DR: A computer procedure to reorganize indexing vocabularies is described and a measure of the semantic association between index terms can be determined from the structural relationships which the terms exhibit by their relative positions in the system.
Abstract: A computer procedure to reorganize indexing vocabularies is described. Index terms are drawn from the vocabulary of a structured indexing system and may consist of single words, collections of words, or syntactic phrases. The basic idea is that a measure of the semantic association between index terms can be determined from the structural relationships which the terms exhibit by their relative positions in the system. The association measure, which is based on a priori (preassigned) semantic relationships between terms, rather than their co-occurrence in a document corpus, is then used for grouping index terms into clusters or concepts. Some results of an experimental investigation are presented. K E Y WORDS AND PHRASES: information, retrieval, clustering,
80 citations
IBM1
TL;DR: An unconventional approach to character recognition is developed, based solely on the statistical properties of the language, therefore it can read printed text with no previous training or a priori information about the structure of the characters.
Abstract: —An unconventional approach to character recognition is developed. The resulting system is based solely on the statistical properties of the language, therefore it can read printed text with no previous training or a priori information about the structure of the characters. The known letter-pair frequencies of the language are used to identify the printed symbols in the following manner.
79 citations
TL;DR: The measure of category clustering proposed by Bousfleld et al. as discussed by the authors was modified to take into account differences in the distribution of the expected value of number of repetitions.
Abstract: The measure of category clustering proposed by Bousfleld & Bousfleld (1966) was modified to take into account differences in the distribution of the expected value of number of repetitions. The possible effects of using the clustering index without this major modification are discussed.
30 citations
TL;DR: A model is introduced that, in addition to having the ability to match error distributions, greatly simplifies the task of code evaluations and is shown to predict the results of tests of an interleaved burst-error-correcting code.
Abstract: The clustering of errors on real channels seriously complicates the task of evaluating the performance of error-control techniques. A model is introduced that, in addition to having the ability to match error distributions, greatly simplifies the task of code evaluations. Having selected the parameters of the model by matching the statistics recorded by Townsend and Watts on the switched telephone network, the model is shown to predict the results of tests of an interleaved burst-error-correcting code. The utility of the model is demonstrated in the analysis of error-detection codes with delayed retransmissions.
24 citations
TL;DR: A document collection consisting of 240 articles on theoretical high energy physics is analyzed by an empirical clustering procedure in which bibliographic coupling, obtained by computer, is used to measure the relatedness of articles.
23 citations
TL;DR: A brief look at the history of the problem outlines three approaches: frequency curve resolution, factor analysis, and hierarchical clustering, based upon distance measures as discussed by the authors, and a generalization for multidimensional distribution resolution is presented.
Abstract: Clustering objects into logical subsets is a frequently encountered problem in biology, psychology, anthropology and many other sciences. Each tends to approach the task in a somewhat different fashion. A brief look at the history of the problem outlines three approaches: frequency curve resolution, factor analysis, and hierarchical clustering, based upon distance measures. A generalization for multidimensional distribution resolution is presented.
20 citations
TL;DR: In this article, the concept of partial weak clustering in the mean is defined for states on physical systems which admit amenable (semi-)groups of symmetries. And the properties of these states are studied and the relations between these states and (extremal) invariant (partial) states are considered.
Abstract: The concept of partial weak clustering in the mean is defined for states on physical systems which admit amenable (semi-)groups of symmetries. The properties of partial weak clustering states are studied and the relations between these states and (extremal) invariant (partial) states are considered. As an application we discuss the zero field magnetization in the two-dimensional Ising model.
5 citations
TL;DR: This paper reported minimal clustering in free recall based on grammatical form-class membership of the items (12 nouns, 12 adjectives, 12 verbs) with zero associative overlap.
4 citations
01 Oct 1968
TL;DR: In this article, the authors suggest some important characteristics of current clustering techniques should be considered in selecting a grouping technique for the data of international relations and suggest that direct factor analysis techniques seem more promising than hierarchical clustering schemes for describing the structure of the spaces defined by the plasmodes.
Abstract: : A research report is made on the applicability of grouping techniques to the data of international relations The results of this paper suggest some important characteristics of current techniques be considered in selecting a grouping technique Several goals guided the research: First, it was desired to demonstrate the characteristics of current clustering techniques by employing experiments with known outcomes--plasmodes; secondly, to suggest a measure of similarity that is advantageous in grouping experiments when correlations are of less meaning; and thirdly, to employ the techniques investigated in the plasmode to a substantive problem from international relations Data were collected on the mileage between a number of cities in the United States The research suggests that direct factor analysis techniques seem more promising than hierarchical clustering schemes for describing the structure of the spaces defined by the plasmodes
3 citations
TL;DR: A method of numerical taxonomy is presented, namely a clustering technique by which bacteria are classified into clusters according to their overall similarity, which is possible to establish a taxonomic system and hence to construct dendrograms.
Abstract: A method of numerical taxonomy is presented, namely a clustering technique by which bacteria are classified into clusters according to their overall similarity. In this way it is possible to establish a taxonomic system and hence to construct dendrograms. It is furthermore possible to evaluate the fundamental characteristics of the clusters by the determination of typical sequences.
01 Mar 1968
TL;DR: A description of various clustering (classification) techniques is provided along with suggestions for alternative approaches and further research.
Abstract: : A description of various clustering (classification) techniques is provided along with suggestions for alternative approaches and further research. (Author)
TL;DR: This article showed that specificity or generality of association is an important determinant of clustering and several other correlations were presented which indicate some possible leads toward research supporting an associationistic interpretation of verbal organization.
Abstract: Clustering scores from an earlier study on short-term recall were correlated with un association test. The results indicated that specificity or generality of association is an important determinant of clustering. Several other correlations were presented which indicate some possible leads toward research supporting an associationistic interpretation of verbal organization.
01 Dec 1968
TL;DR: A sequential algorithm for designing piecewise linear classification functions without a priori knowledge of pattern class distributions is described, which combines, under control of a performance criterion, adaptive error correcting linear classifier design procedures and clustering techniques.
Abstract: A sequential algorithm for designing piecewise linear classification functions without a priori knowledge of pattern class distributions is described. The algorithm combines, under control of a performance criterion, adaptive error correcting linear classifier design procedures and clustering techniques. An error rate criterion is used to constrain the classification function structure so as to minimize design calculations and to increase recognition throughput for many classification problems. Examples from the literature are used to evaluate this approach relative to other classification algorithms.