a comparison of mcc and cen error measures in multi-class prediction比较mcc和岑误差措施,多层次的预测.pdfVIP

A Comparison of MCC and CEN Error Measures in Multi- Class Prediction Giuseppe Jurman*, Samantha Riccadonna, Cesare Furlanello Fondazione Bruno Kessler, Trento, Italy Abstract 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. Citation: Jurman G, Riccadonna S, Furlanello C (2012) A Comparison of MCC and CEN Error Measures in Multi-Class Prediction. PLoS ONE 7(8): e41882. doi:10.1371/journal.pone.0041882 Editor: Giuseppe Biondi-Zoccai, Sapienza University of Rome, Italy Received January 31, 2012; Accepted June 27, 2012; Published August 8, 2012 Copyright: 2012 Jurman et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: The authors acknowledge funding by the European Union FP7 Project HiperDART and by the PAT funded Project ENVIROCHANGE. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: jurman@fbk.eu Introduction For binary tasks, MCC has attracted the attention of the machine learning community as a method that summarizes into a Comparing clas