mis_rate: Classification Error for Heteroscedastic Gaussian Models...

In ywwry66/QDA-by-Projection-R-Package: QDA by Projection

Classification Error for Heteroscedastic Gaussian Models after 1-D Projection

Description

This function computes the population classification error/misclassification rate for heteroscedastic Gaussian models after 1-D projection, given means and covariances of the model and the direction of projection.

Usage

mis_rate(a, mu0, mu1, sigma0, sigma1, p0, p1)

Arguments

a	A vector representing the direction of projection.
mu0	A vector representing the mean of class 0
mu1	A vector representing the mean of class 1
sigma0	An intertible matrix representing the covariance matrix of class 0
sigma1	An invertible matrix representing the covariance matrix of class 1

Details

This function only handles two-class heteroscedastic Gaussian models.

Value

The return value is the classification error/misclassification rate of the two-class heteroscedastic Gaussian model, whose class means are 'mu0' and 'mu1' and class covariance matrices are 'sigma0' and 'sigma1', after 1-D projection onto the direction 'a'.

ywwry66/QDA-by-Projection-R-Package documentation built on Nov. 22, 2023, 2:33 a.m.