Approximating Linear Programs with Summary Functions: Pseudodata with an Infinite Sample
提出一种利用线性规划响应特性来汇总模型的新算法，等价于基于无限样本的伪数据方法，通过顺序采样确定样本点位置，并与基于样本外预测的伪数据方法比较，显示出明显改进。
Abstract A new algorithm for summarizing linear program models that takes advantage of the nature of linear program responses is presented. This method is equivalent to the pseudodata method based on an infinite sample. The power of the new method lies in the fact that the sampling procedure is sequential; the location of the next sample point is determined by observations at previous sample points. An illustration is provided, and the proposed method is compared with pseudodata based on out‐of‐sample predictions. The comparison indicates the new method is a distinct improvement.