A cornerstone of operations management study and practice is the newsvendor problem. This problem began as a simple one-period problem for a price-taking newspaper salesman. Since then, it has motivated several researchers and improved inventory control at numerous companies. The expansion of the choice variables to include price was an important extension of the newsvendor problem. This extension has been known for decades but recently gained a new story since researchers have expended considerable energy in analyzing and exploring this problem under data-driven decision-making. We investigate data-driven approaches for pricing decisions of a retailer where there is only one perishable product with price-dependent stochastic demand. As well-known in marketing, the attraction model is adopted to quantify customer behavior. We also use real retailing data to study the pricing decision of a retailer under the uncertainty of its inventory record inaccuracy.