Page 145 - Data Science Algorithms in a Week

Clustering into K Clusters

th
th
th
Now the 15 couple is in the cluster 1, 16 couple in the cluster 2, 17 couple in the cluster 2.
So the estimated number of the children for each couple is 5/4=1.25.
The error E3 of the estimation is:
2
2
2
E3=sqrt((1.25-1) +(1.25-0) +(1.25-3) )~2.17
Output for 4 clusters:
Cluster 0: [(1, (48.0, 49.0)), (4, (49.0, 42.0)), (10, (42.0, 47.0)), (12,
(41.0, 45.0))]
Cluster 1: [(3, (24.0, 28.0)), (6, (24.0, 27.0)), (11, (22.0, 27.0))]
Cluster 2: [(2, (40.0, 43.0)), (13, (39.0, 43.0)), (14, (36.0, 38.0)), (16,
(36.0, 38.0)), (17, (36.0, 39.0)), (18, (37.0, 38.0))]
Cluster 3: [(5, (32.0, 34.0)), (7, (29.0, 32.0)), (8, (35.0, 35.0)), (9,
(33.0, 36.0)), (15, (30.0, 32.0))]
th
th
th
The 15 couple is in the cluster 3, 16 in the cluster 2, 17 in the cluster 2. So the estimated
number of the children for the 15 couple is 5/4=1.25. The estimated number of the children
th
th
for the 16 and 17 couple is 8/3~2.67 children.
th
The error E4 of the estimation is:
2
2
E4=sqrt((1.25-1) +(8/3-0) +(8/3-3) )~2.70
2
Output for 5 clusters:
Cluster 0: [(1, (48.0, 49.0)), (4, (49.0, 42.0))]
Cluster 1: [(3, (24.0, 28.0)), (6, (24.0, 27.0)), (11, (22.0, 27.0))]
Cluster 2: [(8, (35.0, 35.0)), (9, (33.0, 36.0)), (14, (36.0, 38.0)), (16,
(36.0, 38.0)), (17, (36.0, 39.0)), (18, (37.0, 38.0))]
Cluster 3: [(5, (32.0, 34.0)), (7, (29.0, 32.0)), (15, (30.0, 32.0))]
Cluster 4: [(2, (40.0, 43.0)), (10, (42.0, 47.0)), (12, (41.0, 45.0)), (13,
(39.0, 43.0))]

The 15 couple is in the cluster 3, 16 in the cluster 2, 17 in the cluster 2. So the estimated
th
th
th
th
number of the children for the 15 couple is 1. The estimated number of the children for the
16 and 17 couple is 5/3~1.67.
th
th
The error E5 of the estimation is:
2
2
2
E5=sqrt((1-1) +(5/3-0) +(5/3-3) )~2.13

[ 133 ]