Managing Limited Retail Space for Basic Products

To evaluate the performance of the heuristics used to compute the solutions under the space sharing and the space dedication strategies, we consider a broad range of parameters and generated 500 random problems instances, each comprising between 5 to 15 candidate product. We then computed the solutions provided by the appropriate algorithms and compared these solutions with the appropriate upper bound solution for each strategy. We define three different criteria for the comparison of performance between space sharing and space dedication strategies. We set the value of the last column to 1 if space sharing is better than dedication, -1 if space dedication is better than sharing, and 0 if the performance is similar. We present the results in three sheets which correspond to the three criteria of comparison.

为评估空间共享（space sharing）与空间专用（space dedication）策略下用于求解的启发式算法（heuristics）性能，我们选取了宽泛的参数范围，共生成500个随机问题实例，每个实例包含5至15个候选产品。随后我们通过适配各策略的算法求解得到对应解，并将这些解与各策略对应的最优上界解进行对比。我们设定了三项性能对比准则，用于评估空间共享与空间专用策略的优劣。我们将结果表格的最后一列取值定义为：若空间共享策略表现更优，则设为1；若空间专用策略表现更优，则设为-1；若二者性能相近，则设为0。最终我们将实验结果整理至三个工作表中，分别对应上述三项对比准则。