20.500.12556/DKUM-89793 Kako poštena so klasifikacijska odločitvena drevesa? How fair are classification decision trees? Poštenost klasifikacijskih odločitvenih dreves je na področju strojnega učenja postala kritično vprašanje. Klasifikacijska in regresijska drevesa (CART) so znana po svoji preprostosti in učinkovitosti pri obravnavanju klasifikacijskih in regresijskih nalog. Vendar lahko ti modeli nehote ohranijo ali celo povečajo pristranskost, prisotno v podatkih, kar vodi do nepoštenih odločitev, ki nesorazmerno prizadenejo določene skupine. To magistrsko delo raziskuje poštenost modelov CART z implementacijo metode FairCART, ki vključuje omejitve poštenosti med postopkom oblikovanja dreves. V delu je ocenjena učinkovitost metode FairCART pri zmanjševanju pristranskosti ob hkratnem ohranjanju kakovosti odločitev, kar omogoča vpogled v kompromise med poštenostjo in točnostjo. Implementacija in rezultati eksperimenta kažejo, da lahko metoda FairCART zmerno zmanjša pristranskost in ohrani splošno kakovost odločitvenega drevesa. The fairness of classification decision trees has become a critical concern in the field of machine learning. Classification And Regression Trees (CART) are renowned for their simplicity and effectiveness in handling classification and regression tasks. However, these models can inadvertently perpetuate or even amplify biases present in the data, leading to unfair decisions that disproportionately affect certain groups. This master's thesis explores the fairness of CART models by implementing the FairCART method, which integrates fairness constraints during the tree-building process. The thesis evaluates the effectiveness of FairCART in reducing biases while maintaining decision quality, providing insights into the trade-offs between fairness and accuracy. The implementation and experimental results demonstrate that the FairCART method can modestly reduce bias and maintain overall decision tree quality. klasifikacijska in regresijska drevesa poštenost v strojnem učenju CART FairCART classification and regression trees fairness in machine learning CART FairCART true true false [A. Kostić] Slovenski jezik Angleški jezik Magistrsko delo/naloga 2024-08-12 15:27:55 2024-09-19 16:16:57 2024-12-16 11:33:05 0000-00-00 00:00:00 2024 0 Maribor 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 004.85(043.2) 219212291 MAG_Kostic_Andrej_2024.pdf MAG_Kostic_Andrej_2024.pdf 1 116569B7D4D8B1A8642082B02DA8B362 e9c4fee1277432ccbc88c4fba7bcff969ec35792362b5e0801df6ed27ae71fc0 cfdc8870-58b2-11ef-a285-00155d000105 https://dk.um.si/Dokument.php?lang=slv&id=182285 Fakulteta za elektrotehniko, računalništvo in informatiko 0 0 0