Scholarlink

Optimizing organizational operations with respect to different decisions and to achieve a variety of objectives has been thoroughly studied. Examples include deciding which transport order to assign to which truck in order to minimize costs or deciding on how much stock to keep of a particular product in order to balance costs and delivery reliability. Business processes bring unique challenges to optimization problems, due to the complexity and uncertainty of the interdependencies between their tasks. The business process optimization competition presented such an optimization problem. Participants had to develop an algorithm that decided when to admit patients for hospital treatment, with various uncertain treatment tasks ahead. Eight different algorithms were submitted and compared to identify the most effective types of algorithm for business process optimization. Four algorithms outperformed the heuristic benchmark, with two of these being reinforcement learning algorithms. This suggests that reinforcement learning is a promising approach for optimizing business processes.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The Business Process Optimization Competition

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Felix Schumann,
  • Matthias Ehrendorfer,
  • Michel Kunkler,
  • Kiran Busch,
  • Henrik Leopold,
  • Leon Urny,
  • Martin Schmauch,
  • Olga Rodzik,
  • Stefanie Lanz,
  • Efe Tıraş,
  • Amgad Al-Zamkan,
  • Jana El Kari,
  • Remco Dijkman

摘要

Optimizing organizational operations with respect to different decisions and to achieve a variety of objectives has been thoroughly studied. Examples include deciding which transport order to assign to which truck in order to minimize costs or deciding on how much stock to keep of a particular product in order to balance costs and delivery reliability. Business processes bring unique challenges to optimization problems, due to the complexity and uncertainty of the interdependencies between their tasks. The business process optimization competition presented such an optimization problem. Participants had to develop an algorithm that decided when to admit patients for hospital treatment, with various uncertain treatment tasks ahead. Eight different algorithms were submitted and compared to identify the most effective types of algorithm for business process optimization. Four algorithms outperformed the heuristic benchmark, with two of these being reinforcement learning algorithms. This suggests that reinforcement learning is a promising approach for optimizing business processes.