5. (50) An owner would like to do a customer segmentation on his online shop based on the money (in $1,000) and time (in hour) the customers spent in a week. He applied a K-means clustering using Euclidean distance with k = 3 on these two variables. Each cluster is represented by a label. The centroid of each cluster is listed in the appendix. < ( a. (10) Can the owner apply classification method for the customer segmentation instead? Explain your answer briefly. < b. (5) Given that there is a new customer and his money spent and time spent are listed in the appendix. Assign him the label by the K-means clustering model. c. (15) An elbow plot is shown in the appendix. Is it reasonable to say that k = 3 for the K-means clustering model? Explain your answer briefly. < d. (10) He interprets that the three labels should be 'normal customer', 'loyal customer' and 'window shopper'. Match each of the three labels with the categories in the K-means clustering model and explain your answer briefly. < e. (10) If he also wants to study the behavior similarity between different customers based on the money and time they spent in a week. Should he apply hierarchical clustering instead of K-means clustering? Explain your answer briefly. 5. Centroid of each cluster: Label = 0 Label = 1 Label = 2 Money spent Time spent 0.1 4.2 3.8 New customer: New customer 1 Elbow Plot: < 3.2 1.2 0.8 Money spent Time spent 1 2 Within group Sum of Square 80000 70000 60000 50000 40000 30000 20000 10000 2 5 Elbow Plot Number of cluster k 7 8 9 10