Internet gaming has become more popular. Players play virtually without being in any physical location. The internet gaming has become an attractive business opportunity where game developers and companies sell game tokens. A business consultant has been commissioned to study the relationship of clusters of virtual games from different locations to the amount of money spent on tokens based on different game sets (level of difficulties). The consultant intends to use a reuse case from this clustering activity as a method to develop unsupervised machine learning. K-means is a data clustering approach for unsupervised machine learning that can separate unlabeled data into a predetermined number of disjoint groups of equal variances – clusters – based on their similarities. Based on the case above and dataset below, develop a K- means cluster. Use 2 means and Euclidean estimates.

Question 1 (a)

From the data provided, design an appropriate graphical display. Comment on the pattern observed.

Question 1(b)

From the list of 6 virtual locations, select 2 groups of virtual gamers randomly.

Question 1(c)

Perform a simple version of market segmentation through Cluster Analysis. Use K-Means cluster analysis with k=2 Hint: use Euclidean distance to compute the distance between data points)
 

Gamel Average Points Purchase Game2 Game3 Game4 Location 1 5.1 3.5 1.4 0.2 Location 2 4.9 3 1.4 0.2 Location 3 7 3.2 4.7 1.4 Location 4 6.4 3.2 4.5 1.5 Location 5 6.3 3.3 6 2.5 Location 6 5.8 2.7 5.1 1.9