Intuitively Assessing ML Model Reliability through Example-Based Explanations and Editing Model Inputs
This work addresses the challenge of making ML model reliability assessment more accessible and intuitive for domain experts like physicians, though it is incremental as it builds on existing interpretability methods.
The paper tackles the problem of assessing machine learning model reliability by introducing two visual analytics modules that use example-based explanations and interactive input editing to make uncertainty more intuitive. In an electrocardiogram classification case study, 14 physicians were better able to align model uncertainty with domain factors compared to a baseline feature importance interface.
Interpretability methods aim to help users build trust in and understand the capabilities of machine learning models. However, existing approaches often rely on abstract, complex visualizations that poorly map to the task at hand or require non-trivial ML expertise to interpret. Here, we present two visual analytics modules that facilitate an intuitive assessment of model reliability. To help users better characterize and reason about a model's uncertainty, we visualize raw and aggregate information about a given input's nearest neighbors. Using an interactive editor, users can manipulate this input in semantically-meaningful ways, determine the effect on the output, and compare against their prior expectations. We evaluate our interface using an electrocardiogram beat classification case study. Compared to a baseline feature importance interface, we find that 14 physicians are better able to align the model's uncertainty with domain-relevant factors and build intuition about its capabilities and limitations.