Exploring Dimensionality Reductions with Forward and Backward Projections
This work addresses the challenge of interpreting dimensionality reductions for data scientists, though it is incremental as it builds on existing visualization methods.
The authors tackled the problem of interpreting dimensionality reductions by introducing forward and backward projection interaction techniques, along with prolines and feasibility map visualizations, integrated into a tool called Praxis. Results from a user study with twelve data scientists showed that these techniques are intuitive and effective for exploratory data analysis and hypothesis generation.
Dimensionality reduction is a common method for analyzing and visualizing high-dimensional data across domains. Dimensionality-reduction algorithms involve complex optimizations and the reduced dimensions computed by these algorithms generally lack clear relation to the initial data dimensions. Therefore, interpreting and reasoning about dimensionality reductions can be difficult. In this work, we introduce two interaction techniques, \textit{forward projection} and \textit{backward projection}, for reasoning dynamically about scatter plots of dimensionally reduced data. We also contribute two related visualization techniques, \textit{prolines} and \textit{feasibility map} to facilitate and enrich the effective use of the proposed interactions, which we integrate in a new tool called \textit{Praxis}. To evaluate our techniques, we first analyze their time and accuracy performance across varying sample and dimension sizes. We then conduct a user study in which twelve data scientists use \textit{Praxis} so as to assess the usefulness of the techniques in performing exploratory data analysis tasks. Results suggest that our visual interactions are intuitive and effective for exploring dimensionality reductions and generating hypotheses about the underlying data.