StandardMeasure: Default Nonconformity measure used in regression...

In conformal: Conformal Prediction for Regression and Classification

Description

Nonconformity measure used by default in the class ConformalRegression to calculate nonconformity scores (alpha). StandardMeasure implements the following nonconformity measure:

alpha = |obs - pred| / pred_error

where alpha is the nonconformity measure, obs is the observed value, pred the predicted value with a point prediction model, and pred_error is the error in prediction predicted with an error model. See conformal for further information about how to derive individual confidence intervals.

Usage

StandardMeasure(obs, pred, error)

Arguments

obs	Observed values for the response variable.
pred	Predicted values for the response variable.
error	Predicted errors calculated with an error model.

Value

A numeric vector with the nonconformity scores (alpha).

Author(s)

Isidro Cortes-Ciriano <[e-mail]>

References

Norinder et al. J. Chem. Inf. Model., 2014, 54 (6), pp 1596-1603 DOI: 10.1021/ci5001168 http://pubs.acs.org/doi/abs/10.1021/ci5001168

See Also

ConformalRegression

conformal documentation built on May 30, 2017, 6:49 a.m.