Market segmentation and cluster analysis
3 important questions on Market segmentation and cluster analysis
What is k-means clustering?
- It is not hierarchical but it is a partitioning technique. Partition is the description of which customer belongs in which cluster, for as many clusters as deemed appropriate.
- Tell the model you want e.g. 3 clusters. For each amount of clusters you have to run separate models.
How does k-means clustering work?
- Tell the computer how many clusters you want.
- The computer randomly assigns each customer to a cluster.
- Then centroids are calculated for each of the variables and for each cluster.
- Distance of customers to centroids is calculated and if the distance of the customer to another centroid is smaller, he/she is added to that cluster.
What are the steps in interpreting the clusters?
- Examining the means of each variable for each segment. Provides a profile as to how the customer segments vary.
- Creating a table or plot of 95% CI intervals for all k segments by all p variables tells which differences are significant. Or ANOVA.
- Compare segments on customer data that were not used to derive clusters.
The question on the page originate from the summary of the following study material:
- A unique study and practice tool
- Never study anything twice again
- Get the grades you hope for
- 100% sure, 100% understanding
Remember faster, study better. Scientifically proven.
