Date: Tue, 2 Feb 1999 02:30:24 -0800 (PST)
From: Hillol Kargupta hillol@eecs.wsu.edu
Subject: Paper on Distibuted Data Mining
Web: http://www.eecs.wsu.edu/~hillol/pubs.html
Title: Collective Data Mining: A New Perspective Toward Distributed Data
Mining
Authors: Hillol Kargupta, Byung-Hoon Park, Daryl Hershberger, and
Erik Johnson
Abstract:
This paper introduces the collective data mining (CDM), a new approach
toward distributed data mining (DDM) from heterogeneous sites. It points
out that naive approaches to distributed data analysis in a heterogeneous
environment may face ambiguous situation and may lead to incorrect global
data model. It also observes that any function can be expressed in a
distributed fashion using a set of appropriate basis functions and
orthonormal basis functions can be effectively used for developing a
general framework for DDM that guarantees correct local analysis,
resulting in desired global data model using minimal data communication.
The paper develops the foundation of CDM, discusses decision tree learning
and polynomial regression in CDM for discrete and continuous variables,
and describes the BODHI, a CDM based experimental system.
Full paper can be obtained from:
http://www.eecs.wsu.edu/~hillol/pubs.html
Copyright © 1999 KDnuggets