Functional Correlations in the Pursuit of Performance Assessment of\n Classifiers

2019/05/12 by Nadezhda Gribkova, Gribkova, Nadezhda, Ričardas Zitikis +1 · 1 citation
Decision Sciences · Mathematics · #Advanced Statistical Methods and Models #Applications (stat.AP) #FOS: Computer and information sciences #FOS: Mathematics #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Methodology (stat.ME) #Multi-Criteria Decision Making #Statistics Theory (math.ST)

paper · pdf · doi:10.48550/arxiv.1905.04667

openalex publication_date 2019/05/12 · openalex created_date 2022/07/29 · openalex updated_date 2026/07/28

Abstract

In statistical classification and machine learning, as well as in social and\nother sciences, a number of measures of association have been proposed for\nassessing and comparing individual classifiers, raters, as well as their\ngroups. In this paper, we introduce, justify, and explore several new measures\nof association, which we call CO-, ANTI- and COANTI-correlation coefficients,\nthat we demonstrate to be powerful tools for classifying confusion matrices. We\nillustrate the performance of these new coefficients using a number of\nexamples, from which we also conclude that the coefficients are new objects in\nthe sense that they differ from those already in the literature.\n

Cited by

• Performance of diverse evaluation metrics in NLP-based assessment and text generation of consumer complaints

Related

• Nominal Association Vector and Matrix
• The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation
• The extension of Pearson correlation coefficient, measuring noise, and selecting features
• Identifying Spurious Correlations using Counterfactual Alignment
• Fair Canonical Correlation Analysis