UnProjection: Leveraging Inverse-Projections for Visual Analytics of High-Dimensional Data

11/02/2021
by   Mateus Espadoto, et al.
7

Projection techniques are often used to visualize high-dimensional data, allowing users to better understand the overall structure of multi-dimensional spaces on a 2D screen. Although many such methods exist, comparably little work has been done on generalizable methods of inverse-projection – the process of mapping the projected points, or more generally, the projection space back to the original high-dimensional space. In this paper we present NNInv, a deep learning technique with the ability to approximate the inverse of any projection or mapping. NNInv learns to reconstruct high-dimensional data from any arbitrary point on a 2D projection space, giving users the ability to interact with the learned high-dimensional representation in a visual analytics system. We provide an analysis of the parameter space of NNInv, and offer guidance in selecting these parameters. We extend validation of the effectiveness of NNInv through a series of quantitative and qualitative analyses. We then demonstrate the method's utility by applying it to three visualization tasks: interactive instance interpolation, classifier agreement, and gradient visualization.

READ FULL TEXT

page 4

page 5

page 6

page 7

page 8

page 9

page 11

page 14

research
09/03/2020

TopoMap: A 0-dimensional Homology Preserving Projection of High-Dimensional Data

Multidimensional Projection is a fundamental tool for high-dimensional d...
research
09/23/2020

Burning sage: Reversing the curse of dimensionality in the visualization of high-dimensional data

In high-dimensional data analysis the curse of dimensionality reasons th...
research
08/04/2014

A Moving Least Squares Based Approach for Contour Visualization of Multi-Dimensional Data

Analysis of high dimensional data is a common task. Often, small multipl...
research
07/12/2019

Improving the Projection of Global Structures in Data through Spanning Trees

The connection of edges in a graph generates a structure that is indepen...
research
04/24/2021

Exploring Multi-dimensional Data via Subset Embedding

Multi-dimensional data exploration is a classic research topic in visual...
research
03/14/2018

Optimal Bounds for Johnson-Lindenstrauss Transformations

In 1984, Johnson and Lindenstrauss proved that any finite set of data in...
research
12/08/2022

SpaceEditing: Integrating Human Knowledge into Deep Neural Networks via Interactive Latent Space Editing

We propose an interactive editing method that allows humans to help deep...

Please sign up or login with your details

Forgot password? Click here to reset