Summary Handbook Statistics Final Exam (CIS) [Distinction Level] Questions and Answers 2023

What is the goal of Cluster Analysis? - -To capture the possible natural groupings in the data -The groupings are called clusters -Unsupervised learning technique -Prototype-based clustering - -Each cluster is represented by a central data object, also called a prototype -the prototype of each cluster is usually the center of the cluster -Density clustering - -A cluster is defined as a dense region where data objects are concentrated. Low density areas can be discarded as noise. Not all records are used because of this -Hierarchical clustering - -Cluster hierarchy is created based on the distance between data points. The output in the form of a tree structure called dendrogram is presented for the analyst to pick the best number of clusters -Model-based clustering - -A cluster is a grouping that has the data points belonging to the same probability distribution -K-Means Clustering - -Is a prototype-based clustering technique, where the data set is divided into k clusters -Cluster Centroid - -Is just an imaginary record containing the column averages -Within-Cluster Variation - -For cluster k is usually calculated as the sum of all pairwise squared Euclidean Distance between observations in the cluster, divided by the total number of observations in the cluster