For each question, tick the correct answer or answers.

Q1) Total inertia, between-class inertia, with-class inertia.
if all classes have the same mean, then the within-class inertia is 0
the between-class inertia is 0 if the means of all the classes are identical
if in a class, all individuals have the same values for all variables, the within-class inertia is 0
the total inertia in a data set does not depend on the number of classes

Q2) Increasing or decreasing the between-class or within-class inertia.
merging two classes increases the between-class inertia
merging two classes increases the within-class inertia
merging two classes decreases the between-class inertia
merging two classes decreases the within-class inertia

Q3) If we use Ward’s method, at each step we join together
the 2 classes which minimize the increase in between-class inertia
the 2 classes which minimize the decrease in between-class inertia
the 2 classes which minimize the increase in within-class inertia
the 2 classes which minimize the decrease in within-class inertia

Q4) Evaluating the quality of a clustering
a good clustering maximizes the between-class inertia for a fixed number of clusters
a good clustering maximizes the within-class inertia for a fixed number of clusters
the optimal number of classes in a clustering is obtained by maximizing the between-class inertia
the total inertia increases as the number of classes increases
by dividing one class into two, we increase the between-class inertia

Q5) Ward’s method
avoids chain effects when constructing a tree
can be used when a Euclidean distance is defined between individuals
gives the same results as the single linkage criterion
gives the same results as the complete linkage criterion

Score =
Correct answers: