A bakery delivers to nine regular customers and wants to split them into delivery zones with K-Means. Their positions on the city grid, in kilometres:
| customer | A | B | C | D | E | F | G | H | I |
|---|---|---|---|---|---|---|---|---|---|
| position |
The bakery ran K-Means for to and got these groups:
| groups | |
|---|---|
| 1 | {A, B, C, D, E, F, G, H, I} |
| 2 | {A, B, C, D, E, F} and {G, H, I} |
| 3 | {A, B, C}, {D, E, F} and {G, H, I} |
| 4 | {A, B}, {C}, {D, E, F} and {G, H, I} |
To read the elbow, they score each clustering by its within-cluster sum of squares (WCSS): take every customer's squared straight-line distance to the centroid of its own group, and add those up over all nine customers. A group's centroid is the average position of its members.
Their rule for where to stop: keep adding a zone as long as the extra zone lowers the WCSS by at least 10% of the WCSS at . The elbow is the at which they stop adding.
What is the WCSS at the elbow? Give your answer rounded to 2 decimal places.