. . . "RIV/62690094:18450/07:00001864!RIV07-GA0-18450___" . . "978-80-969562-4-1" . "A Comparison of Software for ROC and AUC Estimates"@en . . "[D604722FFEEB]" . . "A Comparison of Software for ROC and AUC Estimates" . "1"^^ . "Receiver Operating Characteristic (ROC) Curve and the Area under the Curve (AUC) are the methods which are frequently used now for visualisation and quantification of the predictive ability of binary classification model. This article reports about the possibilities of some software for AUC estimates. Statistical packages NCSS, SPSS, and MedCalc are compared with the most complete freeware ROCkit. Other possibilities (R, Matlab and web based applications) are mentioned as well. Finally some results of simulation experiments based on bootstrap application are presented. Results of simulation experiment with bootstrap estimates indicated unbiased estimates of AUC and the narrowest confidence limits." . . "Bratislava" . "Slovensk\u00E1 technick\u00E1 univerzita v Bratislave" . "1"^^ . "AUC; ROC Curve; classification; software"@en . "18450" . . "Receiver Operating Characteristic (ROC) Curve and the Area under the Curve (AUC) are the methods which are frequently used now for visualisation and quantification of the predictive ability of binary classification model. This article reports about the possibilities of some software for AUC estimates. Statistical packages NCSS, SPSS, and MedCalc are compared with the most complete freeware ROCkit. Other possibilities (R, Matlab and web based applications) are mentioned as well. Finally some results of simulation experiments based on bootstrap application are presented. Results of simulation experiment with bootstrap estimates indicated unbiased estimates of AUC and the narrowest confidence limits."@en . "407837" . . "Porovn\u00E1n\u00ED software pro odhady ROC a AUC"@cs . . "5"^^ . "Aplimat 2007, 6th International Conference Proceedings" . . . . "A Comparison of Software for ROC and AUC Estimates" . . "A Comparison of Software for ROC and AUC Estimates"@en . . . "K\u0159ivka ROC (Receiver Operating Characteristic) a plocha pod touto k\u0159ivkou AUC (Area Under the Curve) jsou metody vhodn\u00E9 a \u010Dasto vyu\u017E\u00EDvan\u00E9 pro vizualizaci a kvantifikaci prediktivn\u00EDch vlastnost\u00ED bin\u00E1rn\u00EDho klasifika\u010Dn\u00EDho modelu. Tento \u010Dl\u00E1nek se zab\u00FDv\u00E1 mo\u017Enostmi vyu\u017Eit\u00ED dostupn\u00E9ho softwaru pro tyto odhady. Jsou porovn\u00E1v\u00E1ny statistick\u00E9 syst\u00E9my program\u016F NCSS, SPSS a MedCalc s programem freewarov\u00FDm ROCkit. Jsou zm\u00EDn\u011Bny rovn\u011B\u017E vlastnosti dal\u0161\u00EDch programov\u00FDch syst\u00E9m\u016F (R, Matlab a webov\u00E9 aplikace). Jsou prezentov\u00E1ny a porovn\u00E1v\u00E1ny s v\u00FD\u0161e zm\u00EDn\u011Bn\u00FDm softwarem tak\u00E9 v\u00FDsledky simula\u010Dn\u00EDho experimentu, kter\u00E9 poskytuje vlastn\u00ED webov\u00E1 aplikace odhadu, kter\u00E1 je zalo\u017Eena na metod\u011B bootstrap. V\u00FDsledky simula\u010Dn\u00EDho experimentu indikuj\u00ED nevych\u00FDlen\u00FD bootstrap odhad AUC a u\u017E\u0161\u00ED meze spolehlivosti v porovn\u00E1n\u00ED s ostatn\u00EDmi metodami odhadu, dostupn\u00FDmi v porovn\u00E1van\u00FDch softwarov\u00FDch implementac\u00EDch."@cs . "Porovn\u00E1n\u00ED software pro odhady ROC a AUC"@cs . "P(GA402/04/1308)" . "RIV/62690094:18450/07:00001864" . "449-453" . "Skalsk\u00E1, Hana" .