A company’s application must automatically group similar customers and products based on attributes, without labeled outcomes. Which machine learning approach should the company use?

Unsupervised learning is the correct approach because it deals with unlabeled data, aiming to find hidden patterns or structures within the dataset. Grouping similar customers and products without predefined categories (labeled outcomes) is a classic clustering problem, which is a core task in unsupervised learning. Supervised learning requires labeled data with known outcomes for training. Reinforcement learning focuses on an agent learning optimal actions through trial and error in an environment. Semi-supervised learning uses a small amount of labeled data combined with a large amount of unlabeled data, which isn't the primary scenario described.