umu.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Preprocessing perceptrons
Umeå University, Faculty of Science and Technology, Department of Computing Science.
2004 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

Reliable results are crucial when working with medical decision support systems. A decision support system should be reliable but also be interpretable, i.e. able to show how it has inferred its conclusions. In this thesis, the preprocessing perceptron is presented as a simple but effective and efficient analysis method to consider when creating medical decision support systems. The preprocessing perceptron has the simplicity of a perceptron combined with a performance comparable to the multi-layer perceptron.

The research in this thesis has been conducted within the fields of medical informatics and intelligent computing. The original idea of the production line as a tool for a domain expert to extract information, build decision support systems and integrate them in the existing system is described. In the introductory part of the thesis, an introduction to feed-forward neural networks and fuzzy logic is given as a background to work with the preprocessing perceptron. Input to a decision support system is crucial and it is described how to gather a data set, decide how many and what kind of inputs to use. Outliers, errors and missing data are covered as well as normalising of the input. Training is done in a backpropagation-like manner where the division of the data set into a training and a test set can be done in several different ways just as the training itself can have variations. Three major groups of methods to estimate the discriminance effect of the preprocessing perceptron are described and a discussion of the trade-off between complexity and approximation strength are included.

Five papers are presented in this thesis. Case studies are shown where the preprocessing perceptron is compared to multi-layer perceptrons, statistical approaches and other mathematical models. The model is extended to a generalised preprocessing perceptron and the performance of this new model is compared to the traditional feed-forward neural networks. Results concerning the preprocessing layer and its connection to multivariate decision limits are included. The well-known ROC curve is described and introduced fully into the field of computer science as well as the improved curve, the QROC curve. Finally a tutorial to the program trainGPP is presented. It describes how to work with the preprocessing perceptron from the moment when a data file is provided to the moment when a new decision support system is built.

Place, publisher, year, edition, pages
Umeå: Datavetenskap, Umeå universitet , 2004. , 158 p.
Series
Report / UMINF, ISSN 0348-0542 ; 04.10
Keyword [en]
Datalogi, Preprocessing perceptron, Production line, Neural networks, Backpropagation, Fuzzy logic, ROC and QROC curves, Multivariate decision limits
Keyword [sv]
Datalogi
National Category
Computer Science
Research subject
Computing Science
Identifiers
URN: urn:nbn:se:umu:diva-234ISBN: 91-7305-645-6 (print)OAI: oai:DiVA.org:umu-234DiVA: diva2:142768
Public defence
2004-05-14, MA121, MIT-huset, Umeå Universitet, Umeå, 13:15
Opponent
Supervisors
Available from: 2004-04-15 Created: 2004-04-15 Last updated: 2010-08-16Bibliographically approved
List of papers
1. A Comparison of Numerical Risk Computational Techniques in Screening for Down's Syndrome
Open this publication in new window or tab >>A Comparison of Numerical Risk Computational Techniques in Screening for Down's Syndrome
1998 In: Industrial Applications of Neural Networks, 1998, 425-432 p.Chapter in book (Other academic)
Identifiers
urn:nbn:se:umu:diva-3872 (URN)981-02-3175-X (ISBN)
Available from: 2004-04-15 Created: 2004-04-15 Last updated: 2010-06-08Bibliographically approved
2. The Generalised Preprocessing Perceptron for Medical Data Analysis: A Case Study for the Polycystic Ovary Syndrome
Open this publication in new window or tab >>The Generalised Preprocessing Perceptron for Medical Data Analysis: A Case Study for the Polycystic Ovary Syndrome
Show others...
1996 (English)In: Cybernetics and Systems '96: Proceedings of the 13th European Meeting on Cybernetics and Systems Research / [ed] Robert Trappl, 1996, 597-602 p.Conference paper, Published paper (Other academic)
National Category
Computer Science
Identifiers
urn:nbn:se:umu:diva-3873 (URN)
Conference
Cybernetics and Systems '96 (EMCSR '96), Austrian Society for Cybernetic Studies, held at the University of Vienna, Austria, 9-12 April 1996
Available from: 2004-04-15 Created: 2004-04-15 Last updated: 2014-02-11Bibliographically approved
3. Receiver operating characteristic (ROC) analysis: valuating discriminance effects among decision support systems
Open this publication in new window or tab >>Receiver operating characteristic (ROC) analysis: valuating discriminance effects among decision support systems
2001 (English)Report (Other academic)
Abstract [en]

An overview of the usage of Receiver Operating Characteristic (ROC) analysis within medicine and computer science is given. A survey of the theory behind the analysis is given together with a presentation of how to create the ROC curve and different methods to interpret it. Both methods that rely on binormal distributions and methods that rely on distribution free methods have been mentioned.

A way to better know the quality of the measurements of sensitivity and specificity is presented. The quality measures and the QROC curve as is also included together with a discussion about optimal cut-offs and the connection to Bayesian decision theory. Results from earlier experiments and case studies will be used to exemplify the use of the ROC and QROC curves.

Place, publisher, year, edition, pages
Umeå: Dept. of Computing Science, Umeå University, 2001. 26 p.
Series
Report / UMINF, ISSN 0348-0542 ; 18
Identifiers
urn:nbn:se:umu:diva-34848 (URN)
Available from: 2010-06-21 Created: 2010-06-21 Last updated: 2010-07-12
4. Preprocessing perceptrons and multivariate reference values
Open this publication in new window or tab >>Preprocessing perceptrons and multivariate reference values
2009 (English)In: Data mining and medical knowledge management: cases and applications / [ed] Petra Berka, Jan Rauch, and Djamel Abdelkader Zighed, Medical Information Science Reference , 2009, 108-121 p.Chapter in book (Other academic)
Abstract [en]

Classification networks, consisting of preprocessing layers combined with well-known classification networks, are well suited for medical data analysis. Additionally, by adjusting network complexity to corresponding complexity of data, the parameters in the preprocessing network can, in comparison with networks of higher complexity, be more precisely understood and also effectively utilised as decision limits. Further, a multivariate approach to preprocessing is shown in many cases to increase correctness rates in classification tasks. Handling network complexity in this way thus leads to efficient parameter estimations as well as useful parameter interpretations.

Place, publisher, year, edition, pages
Medical Information Science Reference, 2009
Identifiers
urn:nbn:se:umu:diva-34617 (URN)10.4018/978-1-60566-218-3.ch005 (DOI)9781605662183,1605662186, e-issn: 9781605662190 (ISBN)
Available from: 2010-06-09 Created: 2010-06-09 Last updated: 2010-08-16Bibliographically approved
5. trainGPP - Users' manual
Open this publication in new window or tab >>trainGPP - Users' manual
2004 (English)Report (Other academic)
Abstract [en]

The generalised preprocessing perceptron (GPP) is a model that has the advantage of few parameters and still good discriminance ability. Still the GPP is very general and the user might need some support when building and training the model. The toolbox trainGPP , written by the author in MATLAB, is described. The toolbox can be provided by the author and is used to create different types of GPP:s and to train them.

In this paper a tutorial of the different parts of trainGPP is given. It is assumed that the reader is familiar with the GPP and the theory behind it. However, an appendix is included where a presentation of the generalised preprocessing perceptron is given together with descriptions of different kind of GPPmodels.

Place, publisher, year, edition, pages
Dept. of Computing Science, Umeå University, 2004. 27 p.
Series
Report / UMINF, ISSN 0348-0542 ; 01
Identifiers
urn:nbn:se:umu:diva-8401 (URN)
Available from: 2008-01-21 Created: 2008-01-21 Last updated: 2010-08-16Bibliographically approved

Open Access in DiVA

fulltext(693 kB)1192 downloads
File information
File name FULLTEXT01.pdfFile size 693 kBChecksum SHA-1
21261d3b2f9aa3dff075773bce43e83645ad986dc1f18ee657036e21d90baaeeca5fdf3a
Type fulltextMimetype application/pdf

Authority records BETA

Kallin Westin, Lena

Search in DiVA

By author/editor
Kallin Westin, Lena
By organisation
Department of Computing Science
Computer Science

Search outside of DiVA

GoogleGoogle Scholar
Total: 1192 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

isbn
urn-nbn

Altmetric score

isbn
urn-nbn
Total: 675 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf