The correct answer is B.

The CPCM course places affinity-type work inside advanced predictive analytics. The official CPCM course material states that advanced category analytics includes ''predictive analytics including collaborative filtering, clustering algorithms, regression models and time-to-event models.'' In category management, affinity modeling is used to identify relationships between items that are bought together. Oracle Retail describes market basket/affinity analysis as using data-mining techniques to search for sales patterns between products within transactions, such as rules connecting products purchased together.

Option B is therefore the best answer because chips and salsa is a classic co-purchase relationship. Option A describes clustering, not affinity modeling. Option C describes switching or substitution analysis. Option D describes sales forecasting, usually handled through regression, time-series, or other forecasting models.