Authors:
Ireneusz Czarnowski
and
Piotr Jędrzejowicz
Affiliation:
Gdynia Maritime University, Poland
Keyword(s):
Distributed data mining, Distributed learning classifiers, Data reduction, Agent-based approach.
Related
Ontology
Subjects/Areas/Topics:
Agents
;
Artificial Intelligence
;
Artificial Intelligence and Decision Support Systems
;
Computational Intelligence
;
Data Mining
;
Databases and Information Systems Integration
;
Distributed and Mobile Software Systems
;
Distributed Problem Solving
;
Enterprise Information Systems
;
Evolutionary Computing
;
Knowledge Discovery and Information Retrieval
;
Knowledge Engineering and Ontology Development
;
Knowledge-Based Systems
;
Machine Learning
;
Multi-Agent Systems
;
Sensor Networks
;
Signal Processing
;
Soft Computing
;
Software Engineering
;
Symbolic Systems
Abstract:
The paper presents an approach to learning classifiers from distributed data, based on a data reduction at a local level. In such case, the aim of data reduction is to obtain a compact representation of distributed data repositories, that include non-redundant information in the form of so-called prototypes. In the paper data reduction is carried out by simultaneously selecting instances and features, finally producing prototypes which do not have to be homogenous and can include different sets of features.
From these prototypes the global classifier based on a feature voting is constructed. To evaluate and compare the proposed approach computational experiment was carried out. The experiment results indicate that data reduction at the local level and next merger of prototypes into the global classifier can produce very good classification results.