Systematic Exploration of Low-Dimensional Projections

A playful analogy may help clarify our goals (Seo and Shneiderman, 2004). Imagine you are dropped by parachute into an unfamiliar place— it could be a forest, prairie, or mountainous area. You could set out in a random direction to see what is nearby and then decide where to turn next. Or you might go toward peaks or valleys. You might 

Abiding by these principles, the rank-by-feature framework has an interface for 1D projections and a separate one for 2D projections. Users can begin their exploration with the main graphical display— histograms for ID and scatterplots for 2D—and they can also study numerical summaries for more details. 

For example, assume that users analyze the U.S. counties data set with 17 demographic, health, and econontic statistics available for each county. The dataset can be thought of as a 17-dimensional dataset. Users can choose Pearson correlation coefficient as a ranking criterion at the rank-by-feature framework if they are interested in linear relationships between dimensions. Then, the rank-by-feature framework calculates scores (in this case, Pearson correlation coefficient) for aU 

The rank-by-feature framework was shown effective in facUitating users exploration of large multidimensional datasets fi om several different research fields including microarray data analysis (Seo and Shneiderman, 2006). We believe that it can be successfully apphed to biomedical datasets that are very often large multidimensional datasets. In the following sections, we introduce visual interface frameworks and ranking criteria for ID and 2D projections for multidimensional datasets. 

---