Comparison the Quality of Classification Algorithms
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F62690094%3A18450%2F05%3A00001786" target="_blank" >RIV/62690094:18450/05:00001786 - isvavai.cz</a>
Result on the web
—
DOI - Digital Object Identifier
—
Alternative languages
Result language
angličtina
Original language name
Comparison the Quality of Classification Algorithms
Original language description
This article summarises and describes different measures of the predictive quality(performance) of classification models. Quantitative measures of quality can becomplemented with the visual representation. One of the best known is the ROC(Receiver Operating Characteristic) curve that represents the performance of binaryclassification model within the full range of conditions (costs and class distributions) ofdiscrimination. The duality between ROC and expected costs of a classifier (cost curve) isdescribed in more details here. Cost curve measures the difference in performance of twoclassifiers directly in expected costs (EC). The use of ROC and EC is highly importantwhen asymmetric misclassification costs, imbalanced probabilities of classes or changingconditions occur. A hybrid classifier can be found on the convex hull (ROCCH) of theROC curves of different classifiers. This is potentially the best classification model forany mixture of outside constraints. Lower envelope of cost cur
Czech name
Porovnání vlastností klasifikačních algoritmů
Czech description
klasifikace, míry prediktivní přesnosti, ROC křivka, křivka nákladů
Classification
Type
D - Article in proceedings
CEP classification
BB - Applied statistics, operational research
OECD FORD branch
—
Result continuities
Project
<a href="/en/project/GA402%2F04%2F1308" target="_blank" >GA402/04/1308: Classification models and the assessment of their predictive properties</a><br>
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)
Others
Publication year
2005
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Proceedings of the 23rd International Conference Mathematical Methods in Economics 2005
ISBN
80-7041-535-5
ISSN
—
e-ISSN
—
Number of pages
6
Pages from-to
344-349
Publisher name
Gaudeamus Universita
Place of publication
Hradec Králové
Event location
—
Event date
—
Type of event by nationality
—
UT code for WoS article
—