Position Paper: Towards Comprehensible Predictive Modeling

Predictive modeling applications have a number of concerns. While predictive accuracy is often the primary concern, other issues such as efficiency and robustness are often considered. It is my position that comprehensibility is another concern in predictive modeling applications, and is often an important one. However, without an understanding of comprehensibility in predictive modeling, it is difficult to develop techniques to support it, or to make tradeoffs that balance it with other concerns. In this paper, I will provide thoughts as to understanding comprehensibility across the predictive modeling process. The initial framing of considering who is comprehending, what are they trying to comprehend, and why are they trying to comprehend it allows us to see a range of possible problems and answers to how we might better support predictive modeling applications. These initial thoughts suggest the importance of considering comprehensibility across the predictive modeling process.