Instances for “Product-Based Approximate Linear Programs for Network Revenue Management”
We provide the instances used in the paper “Product-Based Approximate Linear Programs for Network Revenue Management”, by Rui Zhang, Saied Samiedaluie, and Dan Zhang, accepted for publication in Operations Research, 2022. This repository contains the 192 instances used in the paper. All the instances used in the paper are provided in a compressed archive. The accompanying data is contained in the following file: • Instances.zip Description: There is one main folder, which contains 192 instances. There are 96 instances for the hub-and-spoke network and 96 instances for the hotel network. For the hub-and-spoke network, a file’s name starts with “HS”. For the hotel network, a file’s name starts with “Hotel". The number right after “HS” or “Hotel” indicates the number of resources. Then, the number of time periods, the fare scaler, and the load factor should be clear from the file’s name. In each file, the first row has the number of resources I. The second row gives the number of time periods T. Then, row 3 to row (2+I) show the resource-product matrix A. The next T rows, row (2 + I + 1) to row (2 + I + T ), have the arrival probability of each product for each period. Then, we have the profits of the products in the following rows.