Package classifiers :: Module q_classifier_mm

Module q_classifier_mm

Subroutines for q_classifier.py that implement a multi-mu model. That is, one where all the classes share a common covariance matrix.

Functions

count_the_classes(data)
How many examples of each class do we have?

source code

build_classifiers(data, N, modelchoice=M.multi_mu)
Builds a set of quadratic classifiers.

source code

Variables
	HUGE = `1e30`

Imports: math, M, QC, dictops

Function Details

Builds a set of quadratic classifiers. Data is a list of datum_tr . It returns a list: [ (evaluation_result, classifier_description), ... ] .