How to Automate Neural Net Based Learning

Roland Linder; Siegfried J. Pöppl

doi:10.1007/3-540-44596-X_17

How to Automate Neural Net Based Learning

Roland Linder, Siegfried J. Pöppl

Institute of Medical Informatics

Abstract

Although neural networks have many appealing properties, yet there is neither a systematic way how to set up the topology of a neural network nor how to determine its various learning parameters. Thus an expert is needed for fine tuning. If neural network applications should not be realisable only for publications but in real life, fine tuning must become unnecessary. In the present paper an approach is demonstrated fulfilling this demand. Moreover referring to six medical classification and approximation problems of the PROBEN1 benchmark collection this approach will be shown even to outperform fine tuned networks.

Original language	English
Title of host publication	MLDM 2001: Machine Learning and Data Mining in Pattern Recognition
Number of pages	11
Volume	2123
Publisher	Springer Berlin Heidelberg
Publication date	01.01.2001
Pages	206-216
ISBN (Print)	978-3-540-42359-1
ISBN (Electronic)	978-3-540-44596-8
DOIs	https://doi.org/10.1007/3-540-44596-X_17
Publication status	Published - 01.01.2001
Event	2nd International Workshop on Machine Learning and Data Mining in Pattern Recognition - Leipzig, Germany Duration: 25.07.2001 → 27.07.2001 Conference number: 104830

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1007/3-540-44596-X_17

Cite this

@inproceedings{c3a7464ceb3d4bc69741be60bcf3d98f,

title = "How to Automate Neural Net Based Learning",

abstract = "Although neural networks have many appealing properties, yet there is neither a systematic way how to set up the topology of a neural network nor how to determine its various learning parameters. Thus an expert is needed for fine tuning. If neural network applications should not be realisable only for publications but in real life, fine tuning must become unnecessary. In the present paper an approach is demonstrated fulfilling this demand. Moreover referring to six medical classification and approximation problems of the PROBEN1 benchmark collection this approach will be shown even to outperform fine tuned networks.",

author = "Roland Linder and P{\"o}ppl, {Siegfried J.}",

year = "2001",

month = jan,

day = "1",

doi = "10.1007/3-540-44596-X_17",

language = "English",

isbn = "978-3-540-42359-1",

volume = "2123",

series = "Lecture Notes in Computer Science",

publisher = "Springer Berlin Heidelberg",

pages = "206--216",

booktitle = "MLDM 2001: Machine Learning and Data Mining in Pattern Recognition",

note = "2nd International Workshop on Machine Learning and Data Mining in Pattern Recognition <br/>, MLDM 2001 ; Conference date: 25-07-2001 Through 27-07-2001",

}

Linder, R & Pöppl, SJ 2001, How to Automate Neural Net Based Learning. in MLDM 2001: Machine Learning and Data Mining in Pattern Recognition . vol. 2123, Lecture Notes in Computer Science, vol. 2123, Springer Berlin Heidelberg, pp. 206-216, 2nd International Workshop on Machine Learning and Data Mining in Pattern Recognition
, Leipzig, Saxony, Germany, 25.07.01. https://doi.org/10.1007/3-540-44596-X_17

TY - GEN

T1 - How to Automate Neural Net Based Learning

AU - Linder, Roland

AU - Pöppl, Siegfried J.

N1 - Conference code: 104830

PY - 2001/1/1

Y1 - 2001/1/1

N2 - Although neural networks have many appealing properties, yet there is neither a systematic way how to set up the topology of a neural network nor how to determine its various learning parameters. Thus an expert is needed for fine tuning. If neural network applications should not be realisable only for publications but in real life, fine tuning must become unnecessary. In the present paper an approach is demonstrated fulfilling this demand. Moreover referring to six medical classification and approximation problems of the PROBEN1 benchmark collection this approach will be shown even to outperform fine tuned networks.

AB - Although neural networks have many appealing properties, yet there is neither a systematic way how to set up the topology of a neural network nor how to determine its various learning parameters. Thus an expert is needed for fine tuning. If neural network applications should not be realisable only for publications but in real life, fine tuning must become unnecessary. In the present paper an approach is demonstrated fulfilling this demand. Moreover referring to six medical classification and approximation problems of the PROBEN1 benchmark collection this approach will be shown even to outperform fine tuned networks.

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-84899409706&origin=inward&txGid=227ddf60a1a2d6303c17758299e313a1

U2 - 10.1007/3-540-44596-X_17

DO - 10.1007/3-540-44596-X_17

M3 - Conference contribution

SN - 978-3-540-42359-1

VL - 2123

T3 - Lecture Notes in Computer Science

SP - 206

EP - 216

BT - MLDM 2001: Machine Learning and Data Mining in Pattern Recognition

PB - Springer Berlin Heidelberg

T2 - 2nd International Workshop on Machine Learning and Data Mining in Pattern Recognition <br/>

Y2 - 25 July 2001 through 27 July 2001

ER -

How to Automate Neural Net Based Learning

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this