Beyond sigmoids: how to obtain well-calibrated probabilities from binary classifiers with beta calibration (Q1688980)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Beyond sigmoids: how to obtain well-calibrated probabilities from binary classifiers with beta calibration
scientific article

    Statements

    Beyond sigmoids: how to obtain well-calibrated probabilities from binary classifiers with beta calibration (English)
    0 references
    0 references
    0 references
    0 references
    12 January 2018
    0 references
    A predetective model is said to be well-calibrated if its predictions match observed distributions in the data. Often used method of the calibration is a logistic calibration. A new type of calibration like calibration based on the beta distribution is considered. It has been shown that implementation of the beta calibration is easy. Several experiments are used to prove the raison d'etre of the new method.
    0 references
    binary classification
    0 references
    classifier calibration
    0 references
    posterior probabilities
    0 references
    logistic function
    0 references
    sigmoid
    0 references
    beta distribution
    0 references

    Identifiers