Question 1

Amazon uses searches and items purchased to create future product marketing recommendations. Additionally, demographics drive additional potential products to be recommended. To do this, what type of market basket analysis is used?

Accepted Answer

Association Rule is used in market basket analysis to find relationships between variables in large databases, like the patterns in items purchased and searches made by Amazon's customers to recommend future products.

Question 2

Using R, what function is used to view the rules by their lift ratios?

Accepted Answer

The `sort` function is used to order rules by their lift ratios or other metrics when working with association rules in R, especially when using the `arules` package.

Question 3

In cluster analysis, measures are used to form clusters. However, when large data sets are imported into R, sometimes the variables do not share the same format. To overcome this, you standardize the data using the _____ function.

Accepted Answer

The correct function to standardize data in R for cluster analysis is the `scale` function. It normalizes the data to have a mean of 0 and a standard deviation of 1, ensuring that variables are on the same scale.

Question 4

The forming of groups into internally homogeneous groups where each has a unique characteristic, different from other groups, is called cluster analysis.

Accepted Answer

Cluster analysis is a statistical technique used to group objects or cases into clusters that are more similar to each other within each cluster than to those in other clusters, effectively forming internally homogeneous groups.

Question 5

The most commonly used approach for hierarchical clustering is divisive clustering.

Accepted Answer

The most commonly used approach for hierarchical clustering is agglomerative clustering, where clusters are formed starting with individual points and merging them into larger clusters step by step.

Question 6

When using k-means clustering, the number of clusters are specified at the end of the analysis to remove overlapping clusters.

Accepted Answer

In k-means clustering, the number of clusters is specified at the beginning of the analysis, not at the end, to guide the partitioning of the data into k groups.

Question 7

When evaluating large data sets, it is customary to cluster large data sets using the k-means to reduce the computation of measures during each iteration compared to hierarchical clustering methods.

Accepted Answer

K-means clustering is generally preferred for large datasets because it is computationally more efficient than hierarchical clustering methods, especially in terms of reducing the computation required during each iteration. Hierarchical methods, which often require computing and updating distance matrices, can be computationally intensive and less scalable for large datasets.

Question 8

When using R, after the data is imported, set.seed function is used to set the random seed and the k function sets the k parameters to preselect the number of clusters.

Accepted Answer

The `set.seed` function is used to ensure reproducibility in R by setting the seed of R's random number generator, not for setting the number of clusters. The selection of the number of clusters in clustering algorithms (like k-means) is not done through a function named `k` but rather by specifying the number of clusters as a parameter in the clustering function itself (e.g., `kmeans(data, centers = k)` where `k` is the number of clusters).

Question 9

In understanding the association rules, it is best to think of them as an If-Then statement.

Accepted Answer

Association rules in data mining are indeed best understood as If-Then statements, where the "If" part represents the antecedent (conditions) and the "Then" part represents the consequent (outcome or result).

Question 10

Under the association rule, a lift ratio between 0 and 1 indicates a positive association.

Accepted Answer

A lift ratio between 0 and 1 indicates a negative association, meaning that the occurrence of one item negatively affects the occurrence of another in the transaction data.

Question 11

If-Then logical statements are constructed with the If portion being the consequent and the Then being the antecedent.

Accepted Answer

In If-Then logical statements, the "If" portion is the antecedent, and the "Then" portion is the consequent.

Question 12

The Ward's method is the use of a different algorithm to minimize the dissimilarity within clusters by using error sum of squares.

Accepted Answer

Ward's method minimizes the within-cluster variance, or equivalently, the error sum of squares, by choosing the pair of clusters to merge at each step that leads to the minimum increase in total within-cluster variance.

Question 13

A dendrogram allows for a visual inspection of the clustering results.

Accepted Answer

A dendrogram is a tree-like diagram that displays the arrangement of the clusters produced by hierarchical clustering, allowing for visual inspection of the clustering results.

Quiz 11: Unsupervised Data Mining