Evidential calibration of binary SVM classifiers

Evidential calibration of binary SVM classifiers

In machine learning problems, the availability of several classifiers trained on different data or features makes the combination of pattern classifiers of great interest. To combine distinct sources of information, it is necessary to represent the outputs of classifiers in a common space via a transformation called calibration. The most classical way is to use class membership probabilities. H...

Evidential Logistic Regression for Binary SVM Classifier Calibration

The theory of belief functions has been successfully used in many classification tasks. It is especially useful when combining multiple classifiers and when dealing with high uncertainty. Many classification approaches such as k-nearest neighbors, neural network or decision trees have been formulated with belief functions. In this paper, we propose an evidential calibration method that transfor...

Beta calibration: a well-founded and easily implemented improvement on logistic calibration for binary classifiers

For optimal decision making under variable class distributions and misclassification costs a classifier needs to produce well-calibrated estimates of the posterior probability. Isotonic calibration is a powerful non-parametric method that is however prone to overfitting on smaller datasets; hence a parametric method based on the logistic curve is commonly used. While logistic calibration is des...

Likelihood-based Vs Distance-based Evidential Classifiers

Visualization and Interpretation of SVM Classifiers

Many machine learning applications involve modeling sparse high dimensional data. Examples include genomics, brain imaging, time series prediction etc. A common problem in such studies is the understanding of complex data-analytic models, especially nonlinear highdimensional models such as Support Vector Machines (SVM). This paper provides a brief survey of the current techniques for the visual...