Update README.md

Author: r0f1

README.md

@@ -143,11 +143,12 @@ Tutorials - [1](https://www.kaggle.com/residentmario/automated-feature-selection
 [prince](https://github.com/MaxHalford/prince) - Dimensionality reduction, factor analysis (PCA, MCA, CA, FAMD).  
 [sklearn](https://scikit-learn.org/stable/modules/generated/sklearn.manifold.MDS.html) - Multidimensional scaling (MDS).  
 [sklearn](https://scikit-learn.org/stable/modules/generated/sklearn.manifold.TSNE.html) - t-distributed Stochastic Neighbor Embedding (t-SNE), [intro](https://distill.pub/2016/misread-tsne/). Faster implementations: [lvdmaaten](https://lvdmaaten.github.io/tsne/), [MulticoreTSNE](https://github.com/DmitryUlyanov/Multicore-TSNE).  
-[sklearn](http://scikit-learn.org/stable/modules/generated/sklearn.decomposition.TruncatedSVD.html) - Truncated SVD (aka LSA).  
-[mdr](https://github.com/EpistasisLab/scikit-mdr) - Dimensionality reduction, multifactor dimensionality reduction (MDR).  
-[umap](https://github.com/lmcinnes/umap) - Uniform Manifold Approximation and Projection, [talk](https://www.youtube.com/watch?v=nq6iPZVUxZU), [explorer](https://github.com/GrantCuster/umap-explorer), [explanation](https://pair-code.github.io/understanding-umap/).  
 [FIt-SNE](https://github.com/KlugerLab/FIt-SNE) - Fast Fourier Transform-accelerated Interpolation-based t-SNE.  
+[umap](https://github.com/lmcinnes/umap) - Uniform Manifold Approximation and Projection, [talk](https://www.youtube.com/watch?v=nq6iPZVUxZU), [explorer](https://github.com/GrantCuster/umap-explorer), [explanation](https://pair-code.github.io/understanding-umap/).  
+UMAP does not preserve global structure any better than t-SNE when using the same initialization - [paper](https://www.biorxiv.org/content/10.1101/2019.12.19.877522v1).  
 [scikit-tda](https://github.com/scikit-tda/scikit-tda) - Topological Data Analysis, [paper](https://www.nature.com/articles/srep01236), [talk](https://www.youtube.com/watch?v=F2t_ytTLrQ4), [talk](https://www.youtube.com/watch?v=AWoeBzJd7uQ).  
+[mdr](https://github.com/EpistasisLab/scikit-mdr) - Dimensionality reduction, multifactor dimensionality reduction (MDR).  
+[sklearn](http://scikit-learn.org/stable/modules/generated/sklearn.decomposition.TruncatedSVD.html) - Truncated SVD (aka LSA).   
 [ivis](https://github.com/beringresearch/ivis) - Dimensionality reduction using Siamese Networks.  
 [trimap](https://github.com/eamid/trimap) - Dimensionality reduction using triplets.