瓶颈
计算机科学
任务(项目管理)
人工智能
机器学习
特征(语言学)
简单(哲学)
人工神经网络
数据挖掘
工程类
语言学
认识论
哲学
嵌入式系统
系统工程
作者
Kushal Chauhan,Rishabh Tiwari,Jan Freyberg,Pradeep Shenoy,Krishnamurthy Dvijotham
标识
DOI:10.1609/aaai.v37i5.25736
摘要
Concept bottleneck models (CBMs) are interpretable neural networks that first predict labels for human-interpretable concepts relevant to the prediction task, and then predict the final label based on the concept label predictions. We extend CBMs to interactive prediction settings where the model can query a human collaborator for the label to some concepts. We develop an interaction policy that, at prediction time, chooses which concepts to request a label for so as to maximally improve the final prediction. We demonstrate that a simple policy combining concept prediction uncertainty and influence of the concept on the final prediction achieves strong performance and outperforms static approaches as well as active feature acquisition methods proposed in the literature. We show that the interactive CBM can achieve accuracy gains of 5-10% with only 5 interactions over competitive baselines on the Caltech-UCSD Birds, CheXpert and OAI datasets.
科研通智能强力驱动
Strongly Powered by AbleSci AI