Files in this item
|(no description provided)|
|Title:||Conjunctive Conceptual Clustering: A Methodology and Experimentation (Learning)|
|Author(s):||Stepp, Robert Earl, III|
|Department / Program:||Computer Science|
|Degree Granting Institution:||University of Illinois at Urbana-Champaign|
|Abstract:||This thesis describes a machine learning methodology called conjunctive conceptual clustering. The methodology can find conceptual patterns in data as illustrated by three sample problems. In one problem, the method is used to rediscover categories of soybean disease when given a collection of 47 descriptions of diseased soybeans having one of four diseases. In a second problem, the method is used to find categories underlying a collection of blocks-world structures. In a third problem, categories of objects having a more complex structure are determined and contrasted with categories generated by people.
The described method of conjunctive conceptual clustering forms clusters of objects (or situations) not on the basis of a numerical similarity measure but on the basis of the "conceptual cohesiveness" of one object to another. The conceptual cohesiveness between two objects depends on the descriptions of the two objects as well as the descriptions of other nearby objects in the given collection and concepts which are available to describe object groups or object configurations as a whole. From a collection of objects, some background domain knowledge, and a goal or purpose for clustering, conceptual clustering generates a hierarchical classification composed of clusters of objects and corresponding conjunctive-form cluster descriptions (concepts). Conceptual clustering is one paradigm of "learning from observation" in which no teacher guides the learning process.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1984.
|Date Available in IDEALS:||2014-12-15|