Discriminant analysis is based on a Gaussian assumption.

It is relevant to perform a linear regression in this situation.

This clustering result is a possible output of standard kmeans, k=2, applied to this dataset.

For 5 test points x1, x2, x3, x4, x5, a (binary) logistic regression yields the following estimated probabilities (of belonging to class 1) :

What is the predicted label for the red dot using the moving window classification rule for a radius R=3 ?

Hint : the radii of the black circles are 1, 2 and 3.

This table presents for a collection of points the true labels and the labels predicted by some classifier. Compute the probability error of the classifier on this data set. The result will be given in decimal form, with a dot (example : 0.3).

What is the predicted label for the red dot using the nearest-neighbor classification rule for k=4 neighbors ?

Using bagging with 10 classification trees, the following predictions are obtained for the label of an observation x :

It is relevant to use clustering based on mixture models in this situation.

What is the name of the species of an iris whose petal length is 3 and petal width 1.5 ?