K. Schadler and F. Wysotzki. A connectionist approach to structural similarity determination as a basis of clustering, classification and feature detection. In Proc. of the 1st European Symposium on the Principles of Data Mining and Knowledge Discovery, Trondheim 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....relational descriptions as proposed by [8] and [24] for instance. Of very interest to the problems to be solved in bioinformatics are machine learning developments relying on clustering algorithms. Biological objects are systems of high complexity such that relational descriptions have to be used [3, 20]. Recently, first order clustering methods are introduced using distance information only [12, 4] The demands for interpreting the data being generated by genome sequencing are expanding more than ever. Machine learning methods may essentially support the analysis, interpretation, and prediction ....

K. Schadler and F. Wysotzki. A connectionist approach to structural similarity determination as a basis of clustering, classification and feature detection. In J. Komorowski and J. Zytkow, editors, Principles of data mining and knowledge discovery. Proceedings of the 1st EuropeanSymposium, pages 254--264. SpringerVerlag, 1997.

....reflected in the application of clustering methods to graphs. Almost all clustering algorithms mentioned in the literature use pattern representations in terms of feature vectors as input [11] 21] Only few work has been done in devising clustering methods for structures [17] 18] 19] 28] [39], 41] although graph representations have been successfully applied as input to a large number of problems in pattern recognition and computer vision [3] 4] The lack of clustering algorithms working on graphs possibly arises from the fundamental problem of determining the similarity between ....

....distance metric and weighted mean are essential to determine the best matching model and to apply the standard CL rule, respectively. The concept of a metric graph space introduced here is a fundamental approach for the following reasons: 1) It embeds well known concepts as structural similarity [39], 38] and weighted mean [5] within a general mathematical framework. The crux is that many structural similarity measures can be reformulated in terms of a symmetric, positive definite bilinear form such that we can apply the whole theory of metric graph spaces. 2) The concept of a metric graph ....

[Article contains additional citation context not shown here]

K. Schadler and F. Wysotzki. A connectionist approach to structural similarity determination as a basis of clustering, classification and feature detection. In Proc. of the 1st European Symposium on the Principles of Data Mining and Knowledge Discovery, Trondheim 1997.

....property or behavior. Using a graph representation of structured objects, this generalisation is a set of common subgraphs that are believed to cause or to prevent the property or behavior in question. An example 1 For a comparison of graph theoretic methods with ILP see [21] Chapter 6. 32] and [31] discuss some more aspects of the graph vs. logic generalisation. is a certain substructure in a chemical structural formula the presence or absence of which causes some biological activity. Common subgraphs provide an injective mapping between the nodes and relations of two structures to be ....

K. Schadler and F. Wysotzki. A connectionist approach to structural similarity determination as a basis of clustering, classification and feature detection. In Proc. of the 1st European Symposium on the Principles of Data Mining and Knowledge Discovery, LNAI. Springer, to appear 1997.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC