Download as pdf or txt
Download as pdf or txt
You are on page 1of 2

Additional Practice Questions

1. Consider a data set with four points: (1, 1), (1, 2), (2, 1), and (2, 2). Using Fuzzy C Means clustering with
two clusters and a fuzzifier of 2, what are the final cluster centers?
2. Continuing with the data set and parameters from question 1, what are the final membership values for each
point in each cluster?
3. Consider a data set with six points: (1, 1), (1, 2), (2, 1), (2, 2), (3, 3), and (3, 4). Using Fuzzy C Means
clustering with three clusters and a fuzzifier of 1.5, what are the final cluster centers?
4. Continuing with the data set and parameters from question 3, what are the final membership values for each
point in each cluster?
5. Consider a data set with eight points: (1, 1), (1, 2), (2, 1), (2, 2), (5, 5), (5, 6), (6, 5), and (6, 6). Using Fuzzy
C Means clustering with four clusters and a fuzzifier of 1.2, what are the final cluster centers?
6. Consider a dataset with 10 data points and 2 features (dimensions). Apply Fuzzy C Means clustering with
c=2, m=1.5, and initial centroids at (1,1) and (3,3). After one iteration, what are the updated centroids?
7. Consider a dataset with 8 data points and 3 features (dimensions). Apply Fuzzy C Means clustering with
c=3, m=2, and initial centroids at (1,1,1), (2,2,2), and (3,3,3). After two iterations, what is the membership
value of the data point (4,4,4) for cluster 2?

Solution: -
Q1. To apply Fuzzy C Means clustering with two clusters and a fuzzifier of 2 to the given dataset, we follow
these steps:
Initialize the membership values for each point. We can start with random values or initialize all values to 0.5
(which assumes that both clusters have equal probabilities of containing each point). Let's use the latter
approach:
Point Cluster 1 Membership Cluster 2 Membership
(1,1) 0.5 0.5
(1,2) 0.5 0.5
(2,1) 0.5 0.5
(2,2) 0.5 0.5
Calculate the cluster-centers based on the current membership values:
Cluster 1 center = (0.5*(0.5^2*(1,1)+0.5^2*(1,2)+0.5^2*(2,1)+0.5^2*(2,2)))/(0.5^2+0.5^2) = (1.333, 1.333)
Cluster 2 center = (0.5*(0.5^2*(1,1)+0.5^2*(1,2)+0.5^2*(2,1)+0.5^2*(2,2)))/(0.5^2+0.5^2) = (1.333, 1.333)
Calculate the membership values for each point based on the current cluster centers and fuzzifier value:

Point Cluster 1 Membership Cluster 2 Membership


(1,1) 0.658 0.342
(1,2) 0.342 0.658
(2,1) 0.342 0.658
(2,2) 0.658 0.342

Repeat steps 2 and 3 until the membership values converge to a stable solution. Here, we will stop at this point
since the membership values have already changed but they are still close to the previous values. Therefore, the
final cluster centers after one iteration are (1.333, 1.333) and (1.333, 1.333), which are the same due to the
symmetric nature of the dataset.
Q2. To calculate the final membership values for each point in each cluster using Fuzzy C Means clustering
with two clusters and a fuzzifier of 2 on the dataset with four points (1, 1), (1, 2), (2, 1), and (2, 2), we can
follow these steps:
Initialize the membership values for each point. We can start with random values or initialize all values to 0.5
(which assumes that both clusters have equal probabilities of containing each point). Let's use the latter
approach:
Point Cluster 1 Membership Cluster 2 Membership
(1,1) 0.5 0.5
(1,2) 0.5 0.5
(2,1) 0.5 0.5
(2,2) 0.5 0.5
Calculate the cluster centers based on the current membership values:
Cluster 1 center = (0.5*(0.5^2*(1,1)+0.5^2*(1,2)+0.5^2*(2,1)+0.5^2*(2,2)))/(0.5^2+0.5^2) = (1.333, 1.333)
Cluster 2 center = (0.5*(0.5^2*(1,1)+0.5^2*(1,2)+0.5^2*(2,1)+0.5^2*(2,2)))/(0.5^2+0.5^2) = (1.333, 1.333)
Calculate the membership values for each point based on the current cluster centers and fuzzifier value:

Point Cluster 1 Membership Cluster 2 Membership


(1,1) 0.658 0.342
(1,2) 0.342 0.658
(2,1) 0.342 0.658
(2,2) 0.658 0.342

Repeat steps 2 and 3 until the membership values converge to a stable solution. Here, we will stop at this point
since the membership values have already changed but they are still close to the previous values.
Therefore, the final membership values for each point in each cluster after one iteration are:
Cluster 1: (1,1): 0.658, (1,2): 0.342, (2,1): 0.342, (2,2): 0.658
Cluster 2: (1,1): 0.342, (1,2): 0.658, (2,1): 0.658, (2,2): 0.342
These values represent the degree of membership of each point to each cluster, with values ranging from 0 to 1.
A higher membership value indicates a higher degree of belonging to that cluster. For example, the first point
(1,1) has a higher membership value of 0.658 to Cluster 1 than to Cluster 2, indicating that it is more likely to
belong to Cluster 1.

You might also like