K-means Clustering
Download as PPTX, PDF0 likes486 views
The document discusses the K-means clustering algorithm, an unsupervised learning technique that groups unlabeled data points into K clusters based on their similarities. It works by randomly initializing K cluster centroids and then iteratively assigning data points to their nearest centroid and recalculating the centroid positions based on the new assignments until convergence is reached. The document notes that K-means clustering requires specifying the number of clusters K in advance and presents the elbow method as a way to help determine an appropriate value of K for a given dataset.
1 of 20
Downloaded 14 times
More Related Content
What's hot (20)
Ad
Similar to K-means Clustering (20)
Ad
More from Sajib Sen (12)
K-means Clustering
- 2. Clustering Algorithm: Supervised Learning: Training set: {(𝑋11, 𝑋21 , 𝑅𝐸𝐷), (𝑋12, 𝑋22, 𝐵𝐿𝑈𝐸), (𝑋13, 𝑋23, 𝑅𝐸𝐷), (𝑋14 , 𝑋24 , 𝐵𝐿𝑈𝐸), …………… (𝑋1 𝑚, 𝑋2 𝑚, 𝑅𝐸𝐷)}
- 3. Clustering Algorithm: Unsupervised Learning: Training set: {𝑋1, 𝑋2 , 𝑋3 ,……, 𝑋 𝑚 } Where, X={𝑥1, 𝑥2}
- 14. K-means algorithm: Input: -K (number of clusters) -Training set{𝑋1, 𝑋2 , 𝑋3 ,……, 𝑋 𝑚 }Where, X={𝑥1, 𝑥2}
- 15. Randomly initialize K cluster centroids μ1, μ2, … . . , μ 𝑘 ; k<m Repeat { for i= 1 to m: 𝐶 𝑖 ∶= index(1 to K) of cluster centroids close to 𝑋 𝑖 for k= 1 to K: μ 𝑘 := average(mean) of points assigned to cluster k Cluster Assignment Step Move Centroids Step
- 16. K means for non-separated cluster:
- 17. K means for non-separated cluster:
- 18. What is the right value of k?
- 19. Elbow Method
- 20. Lets play with real dataset!!