What type of learning involves discovering patterns in unlabeled data?

Unsupervised learning is the type of learning that focuses on discovering patterns in unlabeled data. In this approach, the system analyzes and identifies structures or groupings within the data without the guidance of predefined labels or categories. This means it can explore the inherent relationships and characteristics of the data, leading to insights such as clustering or identifying hidden structures.

For example, clustering algorithms like K-means or hierarchical clustering are commonly used in unsupervised learning to group similar items together based solely on their features—without knowing beforehand what those groups or categories represent. This capability makes unsupervised learning particularly valuable in fields such as market segmentation, anomaly detection, and data compression.

In contrast, other types of learning, such as supervised learning, rely on labeled data to train models, where each training example is paired with a correct output. Reinforcement learning focuses on learning through trial and error to maximize cumulative rewards in an environment, while transductive learning involves making predictions about a specific set of observations rather than generalizing to new, unseen data.