Variants to a knapsack decomposition heuristic for solving R&D pipeline management problems

详细信息    查看全文

• 作者：Brianna Christian ; Selen Cremaschi ; ^{selen-cremaschi@auburn.edu} ; ^{selen.aydogan@gmail.com}
• 关键词：Multistage stochastic programming ; Pharmaceutical R&D pipeline management ; Decision dependent uncertainty ; Endogenous uncertainty ; Knapsack decomposition algorithm
• 刊名：Computers & Chemical Engineering
• 出版年：2017
• 出版时间：4 January 2017
• 年：2017
• 卷：96
• 期：Complete
• 页码：18-32
• 全文大小：1533 K
• 卷排序：96

文摘

KDA performance is not significantly impacted by problem parameter values. Algorithm time increases superlinearly with increases in problem size. Algorithm time is approximately O(nlog(n)) with respect to knapsack problems. Knapsack generation using AER and ETP improves the quality of the KDA solution. Solution quality diminished using probabilistic and penalty constrained knapsacks.