Accelerating t-SNE using tree-based algorithms

By L. V. D. van der Maaten. 2014

2,649 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Information Systems, Representation & Self-Supervised Learning, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Information Systems

In Representation & Self-Supervised Learning

In Signal Processing

Citations in the corpus

Learning Visual Features from Large Weakly Supervised Data (2015, methodology): “To obtain more insight into the features learned by the models, we applied t-SNE [51, 52] to features extracted from the penultimate layer of an AlexNet trained on 1 , 000 words.”
UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction (2018, methodology): “It should be noted that MulticoreTSNE is a heavily optimized implementation written in C++ based on Van der Maaten’s bhtsne [26] code.”
Interpretable Machine Learning: Fundamental Principles and 10 Grand Challenges (2021, background): “The most famous among them are BH t-SNE [296] and FIt-SNE [187].”
Interpreting Deep Classifier by Visual Distillation of Dark Knowledge (2018, background): “DarkSight also scales better than the faster O ( N log N ) variants of t-SNE (van der Maaten, 2014; Linderman et al., 2017).”
Understanding How Dimension Reduction Tools Work: An Empirical Approach to Deciphering t-SNE, UMAP, TriMAP, and PaCMAP for Data Visualization (2020, methodology): “See van der Maaten and Hinton (2008) for additional details and modifications that were made to the basic algorithm to improve its performance.”
openTSNE: a modular Python library for t-SNE dimensionality reduction and embedding (2019, methodology): “Van Der Maaten (2014) observed that evaluating the attractive forces between all pairs of data points is excessive.”
Visualizing Large-scale and High-dimensional Data (2016, neutral): “Next, we introduce the two components respectively.”
Adversarial Continual Learning (2020, methodology): “Here we illustrate the role of adversarial learning in factorizing the latent space learned for a sequence of tasks in a continual learning setting using the T-distributed Stochastic Neighbor Embedding (T-SNE) [36] plots for the 20-Split miniImageNet experiment.”
Adversarially Regularized Graph Autoencoder (2018, methodology): “We visualize the Cora data in a two-dimensional space by applying the t-SNE algorithm [Van Der Maaten, 2014] on the learned embedding.”
Unsupervised Deep Embedding for Clustering Analysis (2015, background): “The complexity of t-SNE is O(n2), where n is the number of data points, but it can be approximated in O(n log n) (van Der Maaten, 2014).”
Clustering with t-SNE, provably (2017, methodology): “roughly, n.20000). (2) Spectral clustering. The t-SNE algorithm, in this regime, behaves like a spectral clustering algorithm; moreover, this algorithm can be written down explicitly. This allows for (1) the use of theory from spectral clustering to rigorously analyze t-SNE and (2) a fast implementa”
In Defense of the Triplet Loss for Person Re-Identification (2017, background): “Figure 1: A small crop of the Barnes-Hut t-SNE [38] of our learned embeddings for the Market-1501 test-set.”
Variational Prototype Replays for Continual Learning (2019, methodology): “3c provide visualizations of learnt variational means of image samples and variational prototypes from each class by projecting these latent representations into 2D space via the t-sne unsupervised dimension reduction method (Van Der Maaten, 2014).”