Skip to content
Branch: master
Find file History
Pull request Compare This branch is 159 commits ahead, 52 commits behind BlackrockDigital:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
..
Failed to load latest commit information.
README.md
intro_slides_pas_feb_2019.pptx
tsne-tutorial.R

README.md

Introduction to t-SNE - link to meeting slides

Zhihua Chen

February 15, 2019

The hands-on part of the meeting. We will do the following things:

  1. Compute plain and spectral t-SNE entirely within R on the MNIST example dataset

    10,000 images of handwritten digits 0-9, 784 pixels each

  2. Do spectral t-SNE on several active research data sets using an interactive web app

    • Precomputed PCA scores
    • Prepare json object with the PCA scores and sample labels
    • Load json to interact-tsne to view the tsne iterations

Requirements:

  • Recent versions of R and RStudio
  • R packages: Rtsne, irlba, rjson
    • paste into your R session: install.packages(c("Rtsne", "irlba", "rjson"))
  • Meeting files
  • Chrome browser and a laptop with webgl2 support (most laptops should be okay)
  • Download sample t-SNE R code

Misc meeting notes

distill.pub how to use t-SNE effectively

You can’t perform that action at this time.