Automatic training of generalized min-max classifiers
Among fuzzy classifiers, min-max networks have the advantage to be trained in a constructive way, by a simple learning procedure. The classification strategy of Simpson's min-max classifier (1992) consists in covering the training data with hyperboxes constrained to have their boundary surfaces...
Saved in:
Published in | Proceedings Joint 9th IFSA World Congress and 20th NAFIPS International Conference (Cat. No. 01TH8569) pp. 3070 - 3075 vol.5 |
---|---|
Main Authors | , , , |
Format | Conference Proceeding |
Language | English |
Published |
IEEE
2001
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | Among fuzzy classifiers, min-max networks have the advantage to be trained in a constructive way, by a simple learning procedure. The classification strategy of Simpson's min-max classifier (1992) consists in covering the training data with hyperboxes constrained to have their boundary surfaces parallel to the coordinate axes of the chosen reference system. In order to obtain a more accurate data coverage, it is possible to adopt a new classification model which allows to arrange the hyperboxes orientation along any direction of the data space. The training algorithm is based on the ARC/PARC technique, which already yields better performances with respect to the original Simpson's algorithm. Although the most important feature of a classifier is its generalization capability, the effectiveness of a training procedure is strictly related to its automation degree. A low automation degree can be a serious drawback for a classification system, since it can prevent an unskilled user from successfully generate an acceptable model. From this point of view, a learning procedure should not depend on any critical parameter. The automation degree of the new classification system is evaluated in the paper. |
---|---|
AbstractList | Among fuzzy classifiers, min-max networks have the advantage to be trained in a constructive way, by a simple learning procedure. The classification strategy of Simpson's min-max classifier (1992) consists in covering the training data with hyperboxes constrained to have their boundary surfaces parallel to the coordinate axes of the chosen reference system. In order to obtain a more accurate data coverage, it is possible to adopt a new classification model which allows to arrange the hyperboxes orientation along any direction of the data space. The training algorithm is based on the ARC/PARC technique, which already yields better performances with respect to the original Simpson's algorithm. Although the most important feature of a classifier is its generalization capability, the effectiveness of a training procedure is strictly related to its automation degree. A low automation degree can be a serious drawback for a classification system, since it can prevent an unskilled user from successfully generate an acceptable model. From this point of view, a learning procedure should not depend on any critical parameter. The automation degree of the new classification system is evaluated in the paper. |
Author | Mascioli, F.M.F. Rizzi, A. Martinelli, G. Panella, M. |
Author_xml | – sequence: 1 givenname: A. surname: Rizzi fullname: Rizzi, A. organization: Dept. of INFO-COM, La Sapienza Univ., Rome, Italy – sequence: 2 givenname: M. surname: Panella fullname: Panella, M. – sequence: 3 givenname: F.M.F. surname: Mascioli fullname: Mascioli, F.M.F. – sequence: 4 givenname: G. surname: Martinelli fullname: Martinelli, G. |
BookMark | eNotj9FKwzAYhQMqqLMvsKu-QGv-JlmSyzKcDoYKbtcjaf6MSJtKUkF9egvb4fCduw_OPbmOY0RClkBrAKofX9vN9v2jbiiFWnMmQV2RQktF5zI5k9-SIudPOocLAMXuiGi_p3EwU-jKKZkQQzyVoy9PGDGZPvyhK4cQq8H8lF1vcg4-YMoP5MabPmNx2QU5bJ7265dq9_a8Xbe7KgDlU-UtawTnTK8s9955LzVYLY2wyipYIe2ccJyCpA2iYV5ophRyq51zRjPOFmR59gZEPH6lMJj0ezx_Y_8MJEbe |
ContentType | Conference Proceeding |
DBID | 6IE 6IL CBEJK RIE RIL |
DOI | 10.1109/NAFIPS.2001.943718 |
DatabaseName | IEEE Electronic Library (IEL) Conference Proceedings IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume IEEE Xplore All Conference Proceedings IEEE Xplore IEEE Proceedings Order Plans (POP All) 1998-Present |
DatabaseTitleList | |
Database_xml | – sequence: 1 dbid: RIE name: IEEE Electronic Library Online url: https://proxy.k.utb.cz/login?url=https://ieeexplore.ieee.org/ sourceTypes: Publisher |
DeliveryMethod | fulltext_linktorsrc |
Discipline | Computer Science |
EndPage | 3075 vol.5 |
ExternalDocumentID | 943718 |
GroupedDBID | 6IE 6IK 6IL AAJGR AAVQY ALMA_UNASSIGNED_HOLDINGS BEFXN BFFAM BGNUA BKEBE BPEOZ CBEJK OCL RIE RIL |
ID | FETCH-LOGICAL-i104t-fb32544396b4ffdff791b97a5b8b816e0cd5d401702eea3f59388e4b9ddda9343 |
IEDL.DBID | RIE |
ISBN | 9780780370784 0780370783 |
IngestDate | Wed Jun 26 19:23:34 EDT 2024 |
IsPeerReviewed | false |
IsScholarly | false |
Language | English |
LinkModel | DirectLink |
MergedId | FETCHMERGED-LOGICAL-i104t-fb32544396b4ffdff791b97a5b8b816e0cd5d401702eea3f59388e4b9ddda9343 |
ParticipantIDs | ieee_primary_943718 |
PublicationCentury | 2000 |
PublicationDate | 20010000 |
PublicationDateYYYYMMDD | 2001-01-01 |
PublicationDate_xml | – year: 2001 text: 20010000 |
PublicationDecade | 2000 |
PublicationTitle | Proceedings Joint 9th IFSA World Congress and 20th NAFIPS International Conference (Cat. No. 01TH8569) |
PublicationTitleAbbrev | NAFIPS |
PublicationYear | 2001 |
Publisher | IEEE |
Publisher_xml | – name: IEEE |
SSID | ssj0000451183 |
Score | 1.3099467 |
Snippet | Among fuzzy classifiers, min-max networks have the advantage to be trained in a constructive way, by a simple learning procedure. The classification strategy... |
SourceID | ieee |
SourceType | Publisher |
StartPage | 3070 |
SubjectTerms | Classification algorithms Design automation Neural networks Particle measurements Performance evaluation Plasma welding Power system modeling Testing Training data Yield estimation |
Title | Automatic training of generalized min-max classifiers |
URI | https://ieeexplore.ieee.org/document/943718 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjZ3Pa4MwFMfD2tNO3Y-O_SaHXWPVJNUcy1jpBiuFrdBbSczLkFEdm8LoX78kasfGDgMPaiCoib6853ufL0I3KbBY23Ut4SZOCFOgicqUIMpEAlJum7WLdzzOx7Mle1jxVcvZ9rUwAOCTzyBwu_5fvi6z2oXKRoJR-yntoV4iRFOqtQunOEyKnZ3eMU9D6hg2tOXrdMesq5kJxWg-md4vnpx7GAVNrz_UVbxxmQ6aqu0PzyR0OSWvQV2pINv-Ijb-87oP0PC7ig8vdvbpEO1BcYQGnYwDbt_qY8QndVV6civuBCNwafBLA6TOt6DxJi_IRn7izK21c-Pks4doOb17vp2RVk2B5NblqohR1OHIqBgrZow2JhGREonkKlVpNIYw01wzh9OJASQ1XNDUDpUSWmspKKMnqF-UBZwinAGTkoaMqtT2aIR0cr8sS0QseSyAnqEj9xDWbw0wY93c__mfZy_QfpPW5bZL1K_ea7iydr5S136EvwBCvKR0 |
link.rule.ids | 310,311,783,787,792,793,799,4057,4058,27937,55086 |
linkProvider | IEEE |
linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjZ1NS8MwGMeDzoOepnPiuzl4Tdc2aZschzg23cbADXYbSZPIkLWiLcg-vUnaKooHoYe-QGibts9Ln-f3B-CWKhJK49eiSIcJIkJJJFLBkNABUzQyh6XNd0ym8XBBHpbRsuZsu14YpZQrPlOeXXX_8mWeljZV1mMEm0_pLtgzbjWNq2atr4SKBaWY59OF5tTHlmKDa8JOs02arhmf9ab9wWj2ZAPEwKvG_aGv4szLoF31bb87KqGtKnnxykJ46fYXs_GfZ34Iut99fHD2ZaGOwI7KOqDdCDnA-r0-BlG_LHLHboWNZATMNXyukNTrrZJws87Qhn_A1Hrba20FtLtgMbif3w1RraeA1iboKpAW2ALJMIsF0VpqnbBAsIRHggoaxMpPZSSJBeqESnGsI4apmSzBpJScYYJPQCvLM3UKYKoI59gnWFAzombcCv6SNGEhj0Km8Bno2Juweq2QGavq-s__3HsD9ofzyXg1Hk0fL8BBVeRll0vQKt5KdWWsfiGu3Wx_AhHEp78 |
openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=Proceedings+Joint+9th+IFSA+World+Congress+and+20th+NAFIPS+International+Conference+%28Cat.+No.+01TH8569%29&rft.atitle=Automatic+training+of+generalized+min-max+classifiers&rft.au=Rizzi%2C+A.&rft.au=Panella%2C+M.&rft.au=Mascioli%2C+F.M.F.&rft.au=Martinelli%2C+G.&rft.date=2001-01-01&rft.pub=IEEE&rft.isbn=9780780370784&rft.spage=3070&rft.epage=3075+vol.5&rft_id=info:doi/10.1109%2FNAFIPS.2001.943718&rft.externalDocID=943718 |
thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=9780780370784/lc.gif&client=summon&freeimage=true |
thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=9780780370784/mc.gif&client=summon&freeimage=true |
thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=9780780370784/sc.gif&client=summon&freeimage=true |