PloS one, 2012-08, Vol.7 (8), p.e41882
2012
Volltextzugriff (PDF)
Details
- Autor(en) / Beteiligte
- Titel
- A comparison of MCC and CEN error measures in multi-class prediction
- Ist Teil von
- PloS one, 2012-08, Vol.7 (8), p.e41882
- Ort / Verlag
- United States: Public Library of Science
- Erscheinungsjahr
- 2012
- Quelle
- MEDLINE
- Beschreibungen/Notizen
- We show that the Confusion Entropy, a measure of performance in multiclass problems has a strong (monotone) relation with the multiclass generalization of a classical metric, the Matthews Correlation Coefficient. Analytical results are provided for the limit cases of general no-information (n-face dice rolling) of the binary classification. Computational evidence supports the claim in the general case.
- Sprache
- Englisch
- Identifikatoren
- ISSN: 1932-6203eISSN: 1932-6203DOI: 10.1371/journal.pone.0041882Titel-ID: cdi_plos_journals_1326512729
- Format
- –
- Schlagworte
- Accuracy, Algorithms, Area Under Curve, Artificial Intelligence, Bioinformatics, Biology, Classification, Comparative analysis, Computational Biology - methods, Computer applications, Computer Science, Computers, Contingency tables, Correlation analysis, Correlation coefficient, Correlation coefficients, Data mining, Entropy, Error analysis, Expected values, Gene expression, Genotype, Information theory, Machine learning, Mathematics, Models, Statistical, Network topologies, Oligonucleotide Array Sequence Analysis, Probability, Problems, Random variables, Reproducibility of Results, ROC Curve, Software, Statistics as Topic