More Advanced Methods; Cluster Analysis

Question 1

Cluster Analysis is an __________ data analysis tool for organizing observed data into meaningful clusters, based on combinations of variables.

Answer 1

exploratory

Question 2

When to look at grouping (cluster) patterns?

Answer 2

• A PT practitioner would like to group patients according to their attributes in order to better treat them with personalized care plan.
• A PT practitioner would like to classify patients based on their individual health records in order to develop specific management strategies that are appropriate to the patients.

Question 3

What are the 2 types of clusters?

Answer 3

1. ) Hierarchical Clustering

2. ) Non-hierarchical Clustering

Question 4

Hierarchical Clustering:

• A set of nested clusters organized using a hierarchical _____.
• The hierarchical methods produce a set of nested clusters in which each pair of individuals or clusters is progressively nested in a larger cluster until only ____ cluster remains, or all individuals in one group are partitioned step by step.

Answer 4

• tree

- one

Question 5

Non-hierarchical Clustering:

• A group of individuals into non-overlapping subsets (clusters) such that each object is in exactly __ cluster.
• The non-hierarchical methods divide a dataset of n individuals into m clusters.
• ________ clustering is the most commonly used non-hierarchical technique.

Answer 5

• one

- K-mean

Question 6

What are the 3 types of clustering techniques?

Answer 6

• Hierarchical Clustering
• K-Mean Clustering
• Two-Step Clustering

Question 7

__________ clustering is a clustering algorithm where

the clustering is mapped into a hierarchy basing its grouping on the inter-cluster similarities or dissimilarities.

Answer 7

Hierarchical

Question 8

What are the two “types” of Hierarchical Clustering?

Answer 8

• Bottom-Up (agglomerative)

- Top-Down (divisive)

Question 9

What is bottom-up clustering?

Answer 9

Starts with 1 single piece of datum and then merge it with others to form larger groups.

Question 10

What is top-down clustering?

Answer 10

Starts with all in one group and then partition the data step by step using a flat clustering algorithm.

Question 11

What are the steps for bottom-up (agglomerative) clustering?

Answer 11

1. ) Assign each item to a cluster.
2. ) Find the closest (most similar) pair of clusters and merge them into a single cluster, so there is now one cluster less.
3. ) Compute distances (similarities) between the new cluster and each of the old clusters.
4. ) Repeat steps 2 and 3 until all items are clustered into a single cluster of the original sample size.

Question 12

What are the limitations of Hierarchical Clustering?

Answer 12

• Arbitrary decisions- necessary to specify both the distance metric and the linkage criteria without any strong theoretical basis for such decisions.
• Data types- works well with continuous data.
• Misinterpretation of dendrogram- selecting the number of clusters using dendrogram may mislead.

Question 13

_______ clustering is a clustering algorithm where data is classified into K number of clusters. This is the most widely used clustering method. Each individual data is mapped into the cluster with its nearest mean.

Answer 13

K-mean

Question 14

What are the steps for K-Mean Clustering?

Answer 14

1. ) Select K points as the initial centroids.
2. ) Assign points to different centroids based on proximity.
3. ) Re-evaluate the centroid of each group.
4. ) Repeat steps 2 and 3 until the best solutions emerges (the centers are stable).

Question 15

What are the limitations of K-Mean Clustering?

Answer 15

• The researcher chooses the number of clusters.
• More Ks (number of clusters), shorter distance from the centroid.
• As an extreme scenario: When every data point is a centroid, the distance is zero.
• What is the optimal K?

Question 16

__________ clustering is a hybri approach where we run pre-clustering first and then run hierarchical methods.

Answer 16

Two-Step

Question 17

What are the features that differentiate Two-Step Clustering from traditional clustering techniques?

Answer 17

• The ability to create clusters based on both categorical and continuous variables.
• Automatic selection of the number of clusters.
• The ability to analyze large data set efficiently.

Question 18

What are the steps for Two-Step Clustering?

Answer 18

1. ) A sequential approach is used to pre-cluster the cases by condensing the variables (pre-clustering).
2. ) The pre-clusters are statistically merged into the desired number of clusters (clustering).

Question 19

What are the limitations of Two-Step Clustering?

Answer 19

• It can take both continuous and categorical data.
• There is no need to enter the number of clusters a priori because it uses indexes of fit (AIC or BIC) to compare each cluster solution to determine which number of cluster is best.

Question 20

Cluster Quality Validation Index:

• ______________ measures how well an individual data is clustered and it estimates the average distance between clusters.
• ______________ displays a measure of how close each point in one cluster is to points in the neighboring cluster.
• Data with a large silhouette coefficient value of almost 1 means what?
• Data with a small silhouette coefficient value of almost 0 means what?
• Data with a negative silhouette coefficient means what?

Answer 20

• silhouette coefficient
• silhouette plot
• very well clustered
• lies between 2 clusters
• probably placed in wrong cluster

Question 21

Silhouette Values:

• 0.5-1.0 = _____
• 0.2-0.5 = _____
• -1.0-0.2 = ______

Answer 21

• Good
• Fair
• Poor

Question 22

Summary for More Advanced Methods; Cluster Analysis:

• The application of ___________ involves grouping similar cases into homogenous groups (called clusters) when the grouping is not previously known.
• With ________ clustering, the clustering is mapped into a hierarchy basing its grouping on the inter-cluster similarities or dissimilarities.
• With _______ clustering, data is classified into K number of clusters mapping each individual data into the cluster with its nearest mean.
• With ________ clustering, a sequential approach is first used to pre-cluster the cases, and second the pre-clusters are statistically merged into the desired number of clusters.
• Two step clustering may be a better choice over hierarchical or k-mean because the two step clustering can work with _________ data and it is not bound to an arbitrary choice of the number of clusters.

Answer 22

• cluster analysis
• hierarchical
• K-mean
• Two-step
• categorical