Clustering 3 (1 question)

There is at least one mistake.
For example, choice i) should be False.

There is at least one mistake.
For example, choice ii) should be False.

There is at least one mistake.
For example, choice iii) should be True.

There is at least one mistake.
For example, choice iv) should be False.

There is at least one mistake.
For example, choice v) should be True.

There is at least one mistake.
For example, choice vi) should be False.

Correct!

	i)	The quality of the model is measured by the distance of the data points from the centers of the Gaussians.
	ii)	There are configurations, where a point is assigned 100% to one of several Gaussians.
	iii)	The model can represent two clusters that overlap like the two strokes of an ’X’.
	iv)	A model fit with some data points assigned half/half to two different clusters is a bad fit.
	v)	The maximum likelihood algorithm for fitting a Gaussian mixture model can get stuck in a local optimum.
	vi)	It is assumed that all clusters have the same number of points.