Statistical embedding: Beyond principal components

TL;DR

This survey reviews recent advances in high-dimensional and nonlinear data embedding techniques, covering methods from principal curves to topological and network data embeddings, and their applications in visualization.

Contribution

It provides a comprehensive overview of diverse embedding methods, compares algorithmic and statistical approaches, and illustrates techniques with simulated data examples.

Findings

Different embedding methods have unique strengths and limitations.

Topological and network embeddings enable analysis of complex data structures.

Visualization techniques like t-SNE, UMAP, and LargeVis effectively represent high-dimensional data.

Abstract

There has been an intense recent activity in embedding of very high dimensional and nonlinear data structures, much of it in the data science and machine learning literature. We survey this activity in four parts. In the first part we cover nonlinear methods such as principal curves, multidimensional scaling, local linear methods, ISOMAP, graph based methods and diffusion mapping, kernel based methods and random projections. The second part is concerned with topological embedding methods, in particular mapping topological properties into persistence diagrams and the Mapper algorithm. Another type of data sets with a tremendous growth is very high-dimensional network data. The task considered in part three is how to embed such data in a vector space of moderate dimension to make the data amenable to traditional techniques such as cluster and classification techniques. Arguably this is the part where the contrast between algorithmic machine learning methods and statistical modeling, the so-called stochastic block modeling, is at its greatest. In the paper, we discuss the pros and cons for the two approaches. The final part of the survey deals with embedding in $\mathbb{R}^ 2$, i.e. visualization. Three methods are presented: $t$-SNE, UMAP and LargeVis based on methods in parts one, two and three, respectively. The methods are illustrated and compared on two simulated data sets; one consisting of a triplet of noisy Ranunculoid curves, and one consisting of networks of increasing complexity generated with stochastic block models and with two types of nodes.