Fast Fourier Transform-accelerated Interpolation-based t-SNE (FIt-SNE)
t-Stochastic Neighborhood Embedding (t-SNE) is a highly successful method for dimensionality reduction and visualization of high dimensional datasets. A popular implementation of t-SNE uses the Barnes-Hut algorithm to approximate the gradient at each iteration of gradient descent. We modified this implementation as follows:
- Computation of the N-body Simulation: Instead of approximating the N-body simulation using Barnes-Hut, we interpolate onto an equispaced grid and use FFT to perform the convolution, dramatically reducing the time to compute the gradient at each iteration of gradient descent. See this post for some intuition on how it works.
- Computation of Input Similiarities: Instead of computing nearest neighbors using vantage-point trees, we approximate nearest neighbors using the Annoy library. The neighbor lookups are multithreaded to take advantage of machines with multiple cores. Using “near” neighbors as opposed to strictly “nearest” neighbors is faster, but also has a smoothing effect, which can be useful for embedding some datasets (see Linderman et al. (2017)). If subtle detail is required (e.g. in identifying small clusters), then use vantage-point trees (which is also multithreaded in this implementation).
- Early exaggeration: In Linderman and Steinerberger (2017), we showed that appropriately choosing the early exaggeration coefficient can lead to improved embedding of swissrolls and other synthetic datase ts
- Late exaggeration: By increasing the exaggeration coefficient late in the optimization process (e.g. after 800 of 1000 iterations) can improve separation of the clusters
Check out our preprint for more details and some benchmarks.
The only prerequisite is FFTW. FFTW and fitsne can be installed as follows:
conda config --add channels conda-forge #if not already in your channels. Needed for fftw. conda install cython numpy fftw pip install fitsne
And you’re good to go!
Bug reports, feature requests, etc.
If you have any problems with this package, please open an issue on the Github repository.
If you use our software, please cite:
George C. Linderman, Manas Rachh, Jeremy G. Hoskins, Stefan Steinerberger, Yuval Kluger. (2017). Efficient Algorithms for t-distributed Stochastic Neighborhood Embedding. (2017) arXiv:1712.09005 (link)
Our implementation is derived from the Barnes-Hut implementation:
Laurens van der Maaten (2014). Accelerating t-SNE using tree-based algorithms. Journal of Machine Learning Research, 15(1):3221–3245. (link)
Release history Release notifications
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size & hash||File type||Python version||Upload date|
|fitsne-1.0.1.tar.gz (171.6 kB) View hashes||Source||None|