Shortcomings of Visualizations for Human-in-the-Loop Machine Learning

color image of bar graph with zig zagging line on an upward trend

Visualization for Machine Learning or VIS4ML can help developers better design, train, and understand their models. A recent paper available on preprint service ArXiv.org and accepted for publication at IEEE VIS, by a faculty fellow at Stanford HAI, finds that researchers in this space are making inappropriately broad claims about their tools' applicability.