arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2101.09648cs.LGcs.HC
收藏

利用专家一致性改进算法决策支持

Leveraging Expert Consistency to Improve Algorithmic Decision Support

University of Texas at Austin(德克萨斯大学奥斯汀分校);University of Cambridge(剑桥大学);Carnegie Mellon University(卡内基梅隆大学)

Maria De-Arteaga, Vincent Jeanselme, Artur Dubrawski, Alexandra Chouldechova

AI总结:

针对机器学习决策支持中的构建鸿沟问题,提出基于影响函数估计专家一致性并结合标签合并的方法,使模型同时学习专家决策与观测结果,有效缩小鸿沟并提升预测性能。

摘要:

机器学习(ML)正越来越多地被用于支持高风险决策。然而,这里经常存在一种构建鸿沟:即决策任务所关注的构建概念与用作训练ML模型标签的代理变量所捕获的信息之间的差距。因此,ML模型可能无法捕获决策标准的重要维度,从而阻碍其在决策支持中的效用。因此,在为决策支持设计ML系统时,一个基本步骤是在可用代理变量中选择一个目标标签。在本研究中,我们探索将历史专家决策作为一种丰富但也不完美的信息源,它可以与观察到的结果相结合以缩小构建鸿沟。我们认为,管理者和系统设计师可能有兴趣在专家彼此表现出一致性的实例中向专家学习,而在其他情况下则从观察到的结果中学习。我们开发了一种方法论,利用组织信息系统中通常可用的信息来实现这一目标。这涉及两个核心步骤。首先,我们提出了一种基于影响函数的方法,用于在数据中的每个案例由单一专家评估时间接估计专家一致性。其次,我们引入了一种标签合并方法,允许ML模型同时从专家决策和观察到的结果中学习。我们的实证评估使用了临床环境中的模拟以及来自儿童福利领域的真实数据,结果表明,所提出的方法成功地缩小了构建鸿沟,与单独从观察到的结果或专家决策中学习相比,产生了更好的预测性能。

英文摘要:

Machine learning (ML) is increasingly being used to support high-stakes decisions. However, there is frequently a construct gap: a gap between the construct of interest to the decision-making task and what is captured in proxies used as labels to train ML models. As a result, ML models may fail to capture important dimensions of decision criteria, hampering their utility for decision support. Thus, an essential step in the design of ML systems for decision support is selecting a target label among available proxies. In this work, we explore the use of historical expert decisions as a rich -- yet also imperfect -- source of information that can be combined with observed outcomes to narrow the construct gap. We argue that managers and system designers may be interested in learning from experts in instances where they exhibit consistency with each other, while learning from observed outcomes otherwise. We develop a methodology to enable this goal using information that is commonly available in organizational information systems. This involves two core steps. First, we propose an influence function-based methodology to estimate expert consistency indirectly when each case in the data is assessed by a single expert. Second, we introduce a label amalgamation approach that allows ML models to simultaneously learn from expert decisions and observed outcomes. Our empirical evaluation, using simulations in a clinical setting and real-world data from the child welfare domain, indicates that the proposed approach successfully narrows the construct gap, yielding better predictive performance than learning from either observed outcomes or expert decisions alone.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑