A Simple Generalisation of the Area Under the ROC Curve for Multiple Class Classification Problems

@article{Hand2004ASG,
  title={A Simple Generalisation of the Area Under the ROC Curve for Multiple Class Classification Problems},
  author={D. Hand and R. Till},
  journal={Machine Learning},
  year={2004},
  volume={45},
  pages={171-186}
}
  • D. Hand, R. Till
  • Published 2004
  • Mathematics, Computer Science
  • Machine Learning
  • The area under the ROC curve, or the equivalent Gini index, is a widely used measure of performance of supervised classification rules. It has the attractive property that it side-steps the need to specify the costs of the different kinds of misclassification. However, the simple form is only applicable to the case of two classes. We extend the definition to the case of more than two classes by averaging pairwise comparisons. This measure reduces to the standard form in the two class case. We… CONTINUE READING

    Figures, Tables, and Topics from this paper.

    Measuring classifier performance: a coherent alternative to the area under the ROC curve
    • D. Hand
    • Computer Science, Mathematics
    • 2009
    • 672
    • PDF
    Volume under the ROC Surface for Multi-class Problems
    • 165
    • Highly Influenced
    • PDF
    Using AUC and accuracy in evaluating learning algorithms
    • 1,032
    • Highly Influenced
    • PDF
    Learning Decision Trees Using the Area Under the ROC Curve
    • 321
    • PDF
    Improving Accuracy and Cost of Two-class and Multi-class Probabilistic Classifiers Using ROC Curves
    • 140
    • PDF
    Maximizing the area under the ROC curve by pairwise feature combination
    • C. Marroccoa, R. P. W. Duinb, F. Tortorellaa
    • 2008
    • 34
    • PDF
    Multi-class ROC analysis from a multi-objective optimisation perspective
    • 127
    • PDF
    An experimental comparison of performance measures for classification
    • 413
    • PDF
    Ordered multiple-class ROC analysis with continuous measurements.
    • 194
    AUC: A Better Measure than Accuracy in Comparing Learning Algorithms
    • 251
    • Highly Influenced
    • PDF

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 24 REFERENCES
    The meaning and use of the area under a receiver operating characteristic (ROC) curve.
    • 16,393
    • PDF
    The use of the area under the ROC curve in the evaluation of machine learning algorithms
    • 4,027
    • Highly Influential
    • PDF
    Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine.
    • 5,707
    • PDF
    Comparing classifiers when the misallocation costs are uncertain
    • 154
    Analysis and Visualization of Classifier Performance: Comparison under Imprecise Class and Cost Distributions
    • 817
    • PDF
    On fully automatic feature measurement for banded chromosome classification.
    • 191
    Classification and Regression Trees
    • 27,188
    • PDF
    Robust Classification Systems for Imprecise Environments
    • 153
    • PDF