A retail company has a large dataset of customer purchases but no predefined labels.

Unsupervised learning is the correct answer because the dataset does not contain predefined labels or known target outcomes. The AI system is identifying natural patterns in the data and grouping customers with similar purchasing behavior. This type of task is commonly called clustering, which is one of the most common applications of unsupervised learning. Supervised learning is incorrect because there are no labeled examples telling the model which customer belongs to which segment. Reinforcement learning is incorrect because the system is not learning through rewards or penalties. Transfer learning involves reusing knowledge from one trained model for another related task, which is not described here. Semi-supervised learning would involve both labeled and unlabeled data, but this scenario only mentions unlabeled data. Therefore, the correct answer is B. unsupervised learning .

===============