In-class Quiz ISLP Chapter 12

Which of the following statements about clustering is true?

• Clustering method partitions observations into distinct groups
• In a clustering partition, the observations within each group is very similar
• Observations in different partitions are quite different to each other
• All of the above
• None of the above

Which of the following statements is true?

• In unsupervised learning, there is no simple goal of analysis
• Unsupervised learning exercise tends to be very objective
• It is hard to assess the results obtained from unsupervised learning
• All of the above
• None of the above

Which of the following is true about the equation in 12.1?

• This set of features has the lowest variance
• By normalized it is meant that the loadings are constrained
• This set of features has the largest variance
• All of the above
• None of the above

Why the sum of squares of the loadings in equation 12.1 have to be equal to 1?

• Otherwise in a minimization problem, the value of these loadings will be too large
• Otherwise in a maximization problem, the value of these loadings will be too large
• Otherwise in a maximization problem, the value of these loadings will be too small
• All of the above
• None of the above

Which of the following is a use of PCA technique?

• It can produce derived variables for use in supervised learning
• It can serve as a tool for data visualization
• It can serve as a tool for dimension reduction in the data
• All of the above
• None of the above

Which of the statements are NOT true about the following table?

• Crime variables shows the most variation across observations
• Percentage of urban population in a state not as important as crime variables in explaining variation in the data
• Dimension of the data has been reduced to 2 from 4
• All of the above
• None of the above

Consider the following mathematical expression: What does the inner most sum means?

• The sum of squared difference between all the observations across all predictors in a single cluster
• The sum of squared difference between all the observations across all predictors in all the clusters
• The sum of squared difference between the same two observations across all predictors in all the clusters
• The sum of squared difference between the same two observations across all predictors in a single cluster
• None of the above

Consider the following mathematical expression: What does the middle and inner most sum jointly means?

• The sum of squared difference between all the observations across all predictors in a single cluster
• The sum of squared difference between all the observations across all predictors in all the clusters
• The sum of squared difference between the same two observations across all predictors in all the clusters
• The sum of squared difference between the same two observations across all predictors in a single cluster
• None of the above

Consider the following mathematical expression: What does the outer, middle and inner most sum jointly means?

• The sum of squared difference between all the observations across all predictors in a single cluster
• The sum of squared difference between all the observations across all predictors in all the clusters
• The sum of squared difference between the same two observations across all predictors in all the clusters
• The sum of squared difference between the same two observations across all predictors in a single cluster
• None of the above

Consider the following diagram called a dendrogram. This diagram has been derived from a simulated data set of 45 observations. What does each of the leaves in this dendrogram represents?

• A single observation
• A single cluster
• Both of the above
• None of the above