Definition:
An unsupervised learning algorithm that groups data into K clusters based on similarity.
Key Points:
Each cluster has a centroid
Points are assigned to the cluster with the nearest centroid
Iteratively updates centroids until clusters stabilize
Applications:
Customer segmentation
Image compression
Market basket analysis
Advantages:
Simple and fast
Works well for large datasets
Limitations:
Requires predefined K
Sensitive to outliers
Not suitable for non-globular clusters
Take quizzes related to this topic and see where you stand!
Start Quiz Now