Hi @lina ,
Normally to determine the optimal number of clusters in K-Means Clustering, you can utilize the Elbow Method or the Silhouette Method. These methods help in identifying the appropriate number of clusters based on the data distribution. In example code snippet below efficiently implements the Elbow method by downloading it from the following link below
https://www.mathworks.com/matlabcentral/fileexchange/65823-kmeans_opt
for determining the optimal number of clusters in K-Means Clustering. By following this guide, you can adapt and apply this method to various datasets while ensuring accurate clustering results and insightful visualizations.
% Example code snippet
% Load or generate sample data
X = rand(100, 2); % Example data: 100 points in 2D
% Run k-means optimization
[IDX,C,SUMD,K] = kmeans_opt(X);
% Print results
fprintf('Optimal number of clusters: %d\n', K);
fprintf('Centroids:\n');
disp(C);
fprintf('Sum of distances: %.4f\n', SUMD);
% Plotting results
figure;
hold on;
gscatter(X(:,1), X(:,2), IDX);
plot(C(:,1), C(:,2), 'kx', 'MarkerSize', 15, 'LineWidth', 3);
title('K-Means Clustering Results');
xlabel('Feature 1');
ylabel('Feature 2');
legend('Cluster 1', 'Cluster 2', 'Cluster 3', 'Centroids');
hold off;
Please see attached plot
The code snippet above provides a clear example of how to perform K-Means clustering, it randomly generates 100 data points in 2D, calls the custom function kmeans_opt to perform K-Means clustering.Displays the optimal number of clusters, centroids, and sum of distances and visualizes the clustering results with data points colored by cluster and centroids marked. Please let me know if you have any further questions.
