Makine öğrenmesinde 1R algoritması ve ikinci kuralın (2R) oluşturulması

Abstract

0$.ø1(gö5(10(6ø1'(5$/*25ø70$6,9(ø.ø1&ø.85$/,152/8ù7858/0$6,7X÷ED'$/<$1Anahtar Kelimeler:0DNLQHg÷UHQPHVL55Özet:%XoDOÃúPDGDPDNLQH÷UHQPHVLQGHNXOODQÃODQ5DOJRULWPDVÃUQHNELUYHULNÂPHVL Â]HULQGHQ GHWD/Oà RODUDN DQODWÃOPÃúWÃU 7HRULVL DQODWÃODQ DOJRULWPD 3/WKRQSURJUDPODPDGLOLLOHJHOLúWLULOPLúWLU3URJUDPÃQLúOH/LúLYHVD/ÃVDOYHULOHULQNDWHJRULNYHULOHUH QDVÃO GQÂúWÂUÂOGÂ÷ D/Qà YHUL NÂPHVL NXOODQÃODUDN DoÃNODQPÃúWÃU 3URMHQLQJHOLúLPVÂUHFLQGH5DOJRULWPDVÃQÃQDUDPDX]D/ÃJHQLúOHWLOPLúYHLNLQFLNXUDOÃQ5ROXúWXUXOPDVà VD÷ODQPÃúWÃU 5 DOJRULWPDVà ÂUHWLOHQ LON NXUDOà NXOODQDUDN NXUDOGDNLQLWHOLN GÃúÃQGD NDODQ QLWHOLNOHU DUDVÃQGDNL HQ /ÂNVHN IUHNDQVD VDKLS QLWHOLN GH÷HULQLVHoLS LON NXUDOD HNOHU <HGL IDUNOà YHUL NÂPHVL Â]HULQGH WHVW HGLOHQ DOJRULWPDODUÃQVRQXoODUÃQDWH]GHD/UÃFD/HUYHULOPLúWLU2OXúDQNXUDOODUYHKDWDRUDQODUÃQÃNXOODQÃFÃ/DVXQPDNLoLQEDVLWELUDUD/Â]KD]ÃUODQPÃúWÃU

1R ALGORITHM IN MACHINE LEARNING AND FORMING SECONDRULE (2R)7X÷ED'$/<$1Keywords : Machine Learning, 1R, 2R.Abstract: In this study, 1R algorithm that is used in machine learning, is explainedwith using a dataset in details. Algorithm, which is explained theoretically, isimplemented by Python programming language. Progress of program anddiscretization process is clarified with same dataset. Search space of 1R is extendedwith some development and second rule (2R) has improved during the thesis process.2R algorithm uses rules that is derived from 1R, choose most accurate value ofattribute that remains in attribute list. Seven commonly used datasets in machinelearning research is selected to test algorithms and results of tests are presented. Avery simple GUI was designed to see the derived rules and error rates moreexplanatory.