arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1805.10587cs.AI
收藏

预测的语义解释

Semantic Explanations of Predictions

INRIA(法国国家信息与自动化研究所);Accenture Labs(埃森哲实验室);A*STAR Artificial Intelligence Initiative(新加坡科技研究局人工智能计划);Institute for InfoComm Research(信息通信研究院)

Freddy Lecue, Jiewen Wu

摘要:

解释的主要目的是将知识传递给人类。这项工作提出了从机器学习模型的预测构建翔实的解释。受社会科学观察的启发,我们的方法从训练样本中选择了表现出对解释至关重要的特殊特征的数据点,例如,与分类预测相反的数据点和代表模型的数据点。随后,通过使用领域本体从所选数据点导出语义概念。这些概念被过滤和排名,以产生信息丰富的解释,提高人类的理解。我们的方法的主要特点是:(1)关于解释的知识是以本体概念的形式捕获的,(2)解释包括对比证据,除了正常的证据,(3)解释是用户相关的。

英文摘要:

The main objective of explanations is to transmit knowledge to humans. This work proposes to construct informative explanations for predictions made from machine learning models. Motivated by the observations from social sciences, our approach selects data points from the training sample that exhibit special characteristics crucial for explanation, for instance, ones contrastive to the classification prediction and ones representative of the models. Subsequently, semantic concepts are derived from the selected data points through the use of domain ontologies. These concepts are filtered and ranked to produce informative explanations that improves human understanding. The main features of our approach are that (1) knowledge about explanations is captured in the form of ontological concepts, (2) explanations include contrastive evidences in addition to normal evidences, and (3) explanations are user relevant.

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。

↑