Visual Analytics for Dimension Reduction Algorithms
Dimension Reduction (DR) Algorithms are used to analyze multivariate tabular data, such as the often-used car dataset (https://www.kaggle.com/toramky/automobile-dataset). They often aim to find a 2-dimensional projection for p-variate tabular data such that the proximity in 2D space encodes similarity in the original space, ie. they find 2 synthetic dimensions. A popular example is principal component analysis (PCA), and its synthetic dimensions are those with the largest variance. PCA is a linear DR algorithm, as its dimensions are a linear combination of the original dimensions, but non-linear algorithms such as UMAP or t-SNE exist. For many of these one needs to set (hyper-)parameters that influence the result greatly. When people need to explain them, the synthetic dimensions themselves are of interest.
Provide a systematic overview of VA research in dimension reduction. Possible focal points (not exhaustive) are how application prototypes link the result to parameters, or how they support explanation of synthetic dimensions.
D. H. Jeong, C. Ziemkiewicz, B. Fisher, W. Ribarsky, and R. Chang, “iPCA: An Interactive System for PCA-based Visual Analytics,” Computer Graphics Forum, vol. 28, no. 3, pp. 767–774, 2009, doi: 10.1111/j.1467-8659.2009.01475.x.
M. Gleicher, “Explainers: Expert Explorations with Crafted Projections,” IEEE Transactions on Visualization and Computer Graphics, vol. 19, no. 12, pp. 2042–2051, Dec. 2013, doi: 10.1109/TVCG.2013.157.
M. Espadoto, R. M. Martins, A. Kerren, N. S. T. Hirata, and A. C. Telea, “Towards a Quantitative Survey of Dimension Reduction Techniques,” IEEE Transactions on Visualization and Computer Graphics, pp. 1–1, 2019, doi: 10.1109/TVCG.2019.2944182.
L. G. Nonato and M. Aupetit, “Multidimensional Projection for Visual Analytics: Linking Techniques with Distortions, Tasks, and Layout Enrichment,” IEEE Transactions on Visualization and Computer Graphics, vol. 25, no. 8, pp. 2650–2673, Aug. 2019, doi: 10.1109/TVCG.2018.2846735.