k-means clustering

K-means clustering is an unsupervised algorithm that partitions $n$ data points into $k$ distinct clusters, minimizing the within-cluster sum of squares (WCSS) to group similar observations.

This is your go-to centroid-based clustering algorithm: it efficiently partitions an unlabeled dataset into a pre-defined number ($k$) of clusters. The process is iterative: first, assign each data point to the nearest cluster centroid (based on Euclidean distance); second, recalculate the centroid as the mean of all assigned points. The algorithm terminates when centroids stabilize, delivering a local optimum for the WCSS objective function. We use this method for high-stakes tasks like customer segmentation (finding $k$ distinct market groups) and vector quantization for image compression, making it a critical tool in the machine learning arsenal.

https://en.wikipedia.org/wiki/K-means_clustering