Learn++: A classifier independent incremental learning algorithm for supervised neural networks

Robi Polikar; Jeff Byorick; Stefan Krause; Anthony Marino; Michael Moreton

Learn++: A classifier independent incremental learning algorithm for supervised neural networks

Robi Polikar, Jeff Byorick, Stefan Krause, Anthony Marino, Michael Moreton

Electrical Engineering

Research output: Contribution to conference › Paper › peer-review

56 Scopus citations

Abstract

A versatile incremental learning algorithm is introduced for supervised neural network type classifiers. The proposed algorithm, called Learn++, exploits the synergistic expressive power of an ensemble of weak classifiers for learning additional information from new data. Learn++ is capable of learning new classes, without forgetting previously acquired knowledge, even when the previously used data is no longer available. Furthermore, Learn++ is independent of the specific type of the classifier, and adds the incremental learning capability to any supervised neural network classifier.

Original language	English (US)
Pages	1742-1747
Number of pages	6
State	Published - 2002
Event	2002 International Joint Conference on Neural Networks (IJCNN '02) - Honolulu, HI, United States Duration: May 12 2002 → May 17 2002

Other

Other	2002 International Joint Conference on Neural Networks (IJCNN '02)
Country/Territory	United States
City	Honolulu, HI
Period	5/12/02 → 5/17/02

All Science Journal Classification (ASJC) codes

Software
Artificial Intelligence

Cite this

@conference{19966df2d4494f9eb32bdc83e54c5864,

title = "Learn++: A classifier independent incremental learning algorithm for supervised neural networks",

abstract = "A versatile incremental learning algorithm is introduced for supervised neural network type classifiers. The proposed algorithm, called Learn++, exploits the synergistic expressive power of an ensemble of weak classifiers for learning additional information from new data. Learn++ is capable of learning new classes, without forgetting previously acquired knowledge, even when the previously used data is no longer available. Furthermore, Learn++ is independent of the specific type of the classifier, and adds the incremental learning capability to any supervised neural network classifier.",

author = "Robi Polikar and Jeff Byorick and Stefan Krause and Anthony Marino and Michael Moreton",

year = "2002",

language = "English (US)",

pages = "1742--1747",

note = "2002 International Joint Conference on Neural Networks (IJCNN '02) ; Conference date: 12-05-2002 Through 17-05-2002",

}

TY - CONF

T1 - Learn++

T2 - 2002 International Joint Conference on Neural Networks (IJCNN '02)

AU - Polikar, Robi

AU - Byorick, Jeff

AU - Krause, Stefan

AU - Marino, Anthony

AU - Moreton, Michael

PY - 2002

Y1 - 2002

N2 - A versatile incremental learning algorithm is introduced for supervised neural network type classifiers. The proposed algorithm, called Learn++, exploits the synergistic expressive power of an ensemble of weak classifiers for learning additional information from new data. Learn++ is capable of learning new classes, without forgetting previously acquired knowledge, even when the previously used data is no longer available. Furthermore, Learn++ is independent of the specific type of the classifier, and adds the incremental learning capability to any supervised neural network classifier.

AB - A versatile incremental learning algorithm is introduced for supervised neural network type classifiers. The proposed algorithm, called Learn++, exploits the synergistic expressive power of an ensemble of weak classifiers for learning additional information from new data. Learn++ is capable of learning new classes, without forgetting previously acquired knowledge, even when the previously used data is no longer available. Furthermore, Learn++ is independent of the specific type of the classifier, and adds the incremental learning capability to any supervised neural network classifier.

UR - http://www.scopus.com/inward/record.url?scp=0036086746&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036086746&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0036086746

SP - 1742

EP - 1747

Y2 - 12 May 2002 through 17 May 2002

ER -

Learn++: A classifier independent incremental learning algorithm for supervised neural networks

Abstract

Other

All Science Journal Classification (ASJC) codes

Other files and links

Fingerprint

Cite this