This problem does not require the use of data mining software and focuses on knowledge of concepts and basic calculations.
Sum of Withinss Over Number of Clusters
(1)
k=2
This problem does not require the use of data mining software and focuses on knowledge of concepts and basic calculations.
Amanda Boleyn, an entrepreneur who recently sold her start-up for a multi-million-dollar sum, is looking for alternate investments for her newfound fortune. She is considering an investment in wine, similar to how some people invest in rare coins and fine art. To educate herself on the properties of fine wine, she has collected data on 13 different characteristics of 178 wines. Amanda has applied k-means clustering to this data for k = 1,...,10 and generated the following plot of total sums of squared deviations. After analyzing this plot, Amanda generates summaries for k = 2, 3, and 4.
Sum of Withinss Over Number of Clusters
2000 Sum of WithinSS 1000 1500 500
Sum(WithinSS - Diff previous Sum(WithinSS
2
4
6
8
10
Number of Clusters
k=2