Home

Pruning Classifiers in a Distributed Meta-Learning System

Andreas L. Prodromidis; Salvatore Stolfo

Title:
Pruning Classifiers in a Distributed Meta-Learning System
Author(s):
Prodromidis, Andreas L.
Stolfo, Salvatore
Date:
Type:
Technical reports
Department:
Computer Science
Permanent URL:
Series:
Columbia University Computer Science Technical Reports
Part Number:
CUCS-011-98
Publisher:
Department of Computer Science, Columbia University
Publisher Location:
New York
Abstract:
JAM is a powerful and portable agent-based distributed data mining system that employs meta-learning techniques to integrate a number of independent classifiers (models) derived in parallel from independent and (possibly) inherently distributed databases. Although meta-learning promotes scalability and accuracy in a simple and straightforward manner, brute force meta-learning techniques can result in large, redundant, inefficient and some times inaccurate meta-classifier hierarchies. In this paper we explore several methods for evaluating classifiers and composing meta-classifiers, we expose their limitations and we demonstrate that meta-learning combined with certain pruning methods has the potential to achieve similar or even better performance results in a much more cost effective manner.
Subject(s):
Computer science
Item views:
216
Metadata:
text | xml

In Partnership with the Center for Digital Research and Scholarship at Columbia University Libraries/Information Services | Terms of Use