Learning Predictive Checklists with Probabilistic Logic Programming
This addresses the challenge of designing interpretable checklists for predictive tasks in domains such as healthcare, offering a more flexible approach than existing methods, though it is incremental in extending checklist learning to non-Boolean data.
The paper tackles the problem of automatically generating predictive checklists from diverse data modalities like images and time series, which was previously limited to Boolean data, and demonstrates that their method outperforms various explainable machine learning techniques on prediction tasks.
Checklists have been widely recognized as effective tools for completing complex tasks in a systematic manner. Although originally intended for use in procedural tasks, their interpretability and ease of use have led to their adoption for predictive tasks as well, including in clinical settings. However, designing checklists can be challenging, often requiring expert knowledge and manual rule design based on available data. Recent work has attempted to address this issue by using machine learning to automatically generate predictive checklists from data, although these approaches have been limited to Boolean data. We propose a novel method for learning predictive checklists from diverse data modalities, such as images and time series. Our approach relies on probabilistic logic programming, a learning paradigm that enables matching the discrete nature of checklist with continuous-valued data. We propose a regularization technique to tradeoff between the information captured in discrete concepts of continuous data and permit a tunable level of interpretability for the learned checklist concepts. We demonstrate that our method outperforms various explainable machine learning techniques on prediction tasks involving image sequences, time series, and clinical notes.