calc_sens_spec.m File Reference

Detailed Description

Calculates sensitivities and specificities for each class (row of cc)

*  sens(i) = cc(i, i)/sum(cc(i, :))
*  spec(i) = sum_{j, j ~= i} c(j, j) / sum_{j, j ~= i} sum(cc(j, :))
* 

Definition in file calc_sens_spec.m.

Go to the source code of this file.

Functions
function	calc_sens_spec (in cc, in flag_mean)

Function Documentation

function calc_sens_spec	(	in	cc,
		in	flag_mean
	)

Parameters

cc	confusion matrix either in HITS, not percentage
flag_mean	if true, the average of all sensitivities and specificities will be returned (this is useful because the cases when sensitivity or specificity is not defined (like rows or columns of cc containing all zeros) are discounted

Returns

a no_classes X 2 matrix. First column are the sensitivities; second column are the specificities