clustering

Unsupervised machine learning: automatically groups similar data points into distinct clusters for pattern recognition and structure discovery.

Clustering is an unsupervised machine learning technique: it automatically partitions unlabeled data into meaningful groups (clusters) where internal similarity is maximized. We use algorithms like K-Means (a centroid-based method) and DBSCAN (a density-based approach) to execute this. Applications are broad and high-value: from segmenting a customer base for targeted marketing strategies to identifying anomalies in network traffic or financial transactions. It’s a critical technique for exploratory data analysis (EDA) and preprocessing large datasets.

https://scikit-learn.org/stable/modules/clustering.html