Statistics - Quadratic discriminant analysis (QDA)

> (Statistics|Probability|Machine Learning|Data Mining|Data and Knowledge Discovery|Pattern Recognition|Data Science|Data Analysis)

1 - About

When the variances of all X are different in each class, the magic of cancellation doesn't occur because when the variances are different in each class, the quadratic terms don't cancel.

And therefore , the discriminant functions are going to be quadratic functions of X.

Quadratic discriminant analysis uses a different covariance matrix for each class.

Quadratic discriminant analysis is attractive if the number of variables is small.


3 - Discriminant Function

<MATH> \delta_k(x) = - \frac{1}{2} (x - \mu_k)^T \sum^{-1}_k ( x - \mu_k) + log(\pi_k) </MATH>

As there's no cancellation of variances, the discriminant functions now have this distance term that involves <math>\sum_k</math>, which is for the kth class.

There's :

  • a term to do with the prior probability,
  • a determinant term that comes from the covariance matrix.

It gives a curved discriminant boundary.