DimReader: Axis lines that explain non-linear projections

Abstract

Non-linear dimensionality reduction (NDR) methods such as LLE and t-SNE are popular with visualization researchers and experienced data analysts, but present serious problems of interpretation. In this paper, we present DimReader, a technique that recovers readable axes from such techniques. DimReader is based on analyzing infinitesimal perturbations of the dataset with respect to variables of interest. The perturbations define exactly how we want to change each point in the original dataset and we measure the effect that these changes have on the projection. The recovered axes are in direct analogy with the axis lines (grid lines) of traditional scatterplots. We also present methods for discovering perturbations on the input data that change the projection the most. The calculation of the perturbations is efficient and easily integrated into programs written in modern programming languages. We present results of DimReader on a variety of NDR methods and datasets both synthetic and real-life, and show how it can be used to compare different NDR methods. Finally, we discuss limitations of our proposal and situations where further research is needed.

13 Pages, 12 Figures

References in corpus (2)

• The NumPy array: a structure for efficient numerical computation
• Vector Field k-Means: Clustering Trajectories by Fitting Multiple Vector Fields

Cited by in corpus (15)

• t-viSNE: Interactive Assessment and Interpretation of t-SNE Projections
• Genetic Programming for Evolving a Front of Interpretable Models for Data Visualisation
• Uncertainty-Aware Principal Component Analysis
• Visualizing Graph Neural Networks with CorGIE: Corresponding a Graph to Its Embedding
• Unveiling High-dimensional Backstage: A Survey for Reliable Visual Analytics with Dimensionality Reduction
• Implicit Multidimensional Projection of Local Subspaces
• A Critical Analysis of the Usage of Dimensionality Reduction in Four Domains
• Conditional t-SNE: Complementary t-SNE embeddings through factoring out prior information
• HyperNP: Interactive Visual Exploration of Multidimensional Projection Hyperparameters
• Feature Clock: High-Dimensional Effects in Two-Dimensional Plots
• Attribute-based Explanations of Non-Linear Embeddings of High-Dimensional Data
• Orthogonal projections of hypercubes
• Exploring Multi-dimensional Data via Subset Embedding
• MoE-Enhanced Explainable Deep Manifold Transformation for Complex Data Embedding and Visualization
• ExClus: Explainable Clustering on Low-dimensional Data Representations