Variable precision rough set model
Journal of Computer and System Sciences
Automated Refinement of First-Order Horn-Clause Domain Theories
Machine Learning
Tolerance approximation spaces
Fundamenta Informaticae - Special issue: rough sets
Inductive logic programming and knowledge discovery in databases
Advances in knowledge discovery and data mining
Knowledge discovery by application of rough set models
Rough set methods and applications
Rough Sets: Theoretical Aspects of Reasoning about Data
Rough Sets: Theoretical Aspects of Reasoning about Data
Learning First-Order Rules: A Rough Set Approach
Fundamenta Informaticae - International Conference on Soft Computing and Distributed Processing (SCDP'2002)
Similarity-Based Classification in Relational Databases
Fundamenta Informaticae
Satisfiability of Formulas from the Standpoint of Object Classification: The RST Approach
Fundamenta Informaticae - Concurrency Specification and Programming (CS&P)
| Hi-index | 0.00 |
In the paper, we propose an algorithm for classification of complex structured objects. The objects, expressed in a first-order logic (FOL) language, are positive and negative examples of a target relation. In the process of searching for a classification pattern, a similarity measure and some notions of rough set theory are applied. We search for the pattern being a similarity degree of examples and satisfying two conditions: the number of positive examples similar at least to the degree to other positive examples is highest and the number of negative examples similar at least to the degree to positive examples is lowest. The obtained set of similar examples corresponds to the lower or to the upper approximation of a set of all positive examples. The found similarity degree is applied in classification of new examples. An example is classified as positive if it belongs to the approximation computed with respect to the degree, and it is classified as negative, otherwise.