Learning from Observation: Conceptual Clustering

Michalski, Ryszard S.; Stepp, Robert E.

doi:10.1007/978-3-662-12405-5_11

Cited by 247 publications

(105 citation statements)

References 2 publications

Supporting

Mentioning

103

Contrasting

Unclassified

Order By: Relevance

“…This is referred to as the "separate and conquer" (Quinlan, 1990) or "covering" (Michalski & Stepp, 1983) strategy. The basic FOIL procedure learns as tbllows.…”

Section: Methods For Learning Multiple Class Descriptionsmentioning

confidence: 99%

Error reduction through learning multiple descriptions

1996

View full text Add to dashboard Cite

Abstract. Learning multiple descriptions for each class in the data has been shown to reduce generalization error but the amount of error reduction varies greatly from domain to domain. This paper presents a novel empirical analysis that helps to understand this variation. Our hypothesis is that the amount of error reduction is linked to the "degree to which the descriptions for a class make errors in a correlated manner." We present a precise and novel definition for this notion and use twenty-nine data sets to show that the amount of observed error reduction is negatively correlated with the degree to which the descriptions make errors in a correlated manner. We empirically show that it is possible to learn descriptions that make less correlated errors in domains in which many ties in the search evaluation measure (e.g. information gain) are experienced during learning. The paper also presents results that help to understand when and why multiple descriptions are a help (irrelevant attributes) and when they are not as much help (large amounts of class noise).

show abstract

“…This is referred to as the "separate and conquer" (Quinlan, 1990) or "covering" (Michalski & Stepp, 1983) strategy. The basic FOIL procedure learns as tbllows.…”

Section: Methods For Learning Multiple Class Descriptionsmentioning

confidence: 99%

Error reduction through learning multiple descriptions

1996

View full text Add to dashboard Cite

show abstract

“…An important feature of conceptual clustering methods (Michalski & Stepp, 1983) is that any output class, in addition to being characterized by its extensional description (i.e., the set of objects covered by the class), is also characterized by an intensional (conceptual) description. A second feature, while not definitional of conceptual clustering, is nonetheless often desirable, is that output classes (often referred to as concepts) are arranged into a hierarchy based on their generality/specificity.…”

Section: Introductionmentioning

confidence: 99%

A lattice conceptual clustering system and its application to browsing retrieval

1996

View full text Add to dashboard Cite

Abstract. The theory of concept (or Galois) lattices provides a simple and formal approach to conceptual clustering. In this paper we present GALOIS, a system that automates and applies this theory. The algorithm utilized by GALOIS to build a concept lattice is incremental and efficient, each update being done in time at most quadratic in the number of objects in the lattice. Also, the algorithm may incorporate background information into the lattice, and through clustering, extend the scope of the theory. The application we present is concerned with information retrieval via browsing, for which we argue that concept lattices may represent major support structures. We describe a prototype user interface for browsing through the concept lattice of a document-term relation, possibly enriched with a thesaurus of terms. An experimental evaluation of the system performed on a medium-sized bibliographic database shows good retrieval performance and a significant improvement after the introduction of background knowledge.

show abstract

“…Several clustering algorithms that generate concept descriptions can be found in the literature. On the one hand we have those coming from traditional conceptual clustering such as CLUSTER/2 [4], COBWEB [7] and GCF [8]. On the other hand we have those that, using a subset of first-order logic as representation language, apply traditional distance-based clustering algorithms.…”

Section: Discussionmentioning

confidence: 99%

“…A different approach to clustering is conceptual clustering defined by Michalski [3,4]. Conceptual clustering overcomes the cluster interpretation problem by forming clusters that can be described by properties involving relations on a selected set of attributes.…”

Section: Introductionmentioning

confidence: 99%

Hierarchical Distance-Based Conceptual Clustering

Funes

Ferri

Hernández-Orallo

et al.

Machine Learning and Knowledge Discovery in Databases

View full text Add to dashboard Cite

Abstract. In this work we analyse the relation between hierarchical distancebased clustering and the concepts that can be obtained from the hierarchy by generalisation. Many inconsistencies may arise, because the distance and the conceptual generalisation operator are usually incompatible. To overcome this, we propose an algorithm which integrates distance-based and conceptual clustering. The new dendrograms can show when an element has been integrated to the cluster because it is near in the metric space or because it is covered by the concept. In this way, the new clustering can differ from the original one but the metric traceability is clear. We introduce three different levels of agreement between the clustering hierarchy obtained from the linkage distance and the new hierarchy, and we define properties these generalisation operators should satisfy in order to produce distance-consistent dendrograms.

show abstract

Learning from Observation: Conceptual Clustering

Cited by 247 publications

References 2 publications

Error reduction through learning multiple descriptions

Error reduction through learning multiple descriptions

A lattice conceptual clustering system and its application to browsing retrieval

Hierarchical Distance-Based Conceptual Clustering

Contact Info

Product

Resources

About