Self-Supervised Noise Embeddings (Self-SNE) for dimensionality reduction and clustering

Self-SNE is a probabilistic self-supervised deep learning model for compressing high-dimensional data to a low-dimensional embedding. It is a general-purpose algorithm that works with multiple types of data including images, sequences, and tabular data. It uses self-supervised objectives, such as InfoNCE, to preserve structure in the compressed latent space. Self-SNE can also (optionally) simultaneously learn a cluster distribution (a prior over the latent embedding) during optimization. Overlapping clusters are automatically combined by optimizing a variational upper bound on entropy, so the number of clusters does not have to be specified manually — provided the number of initial clusters is large enough. Self-SNE produces embeddings with similar quality to existing dimensionality reduction methods; can detect outliers; scales to large, out-of-core datasets; and can easily add new data to an existing embedding/clustering.

This is an alpha release currently undergoing development. Features may change without notice. Use at your own risk.

References

If you use Self-SNE for your research please cite version 1 of our preprint (an updated version is forthcoming):

@article{graving2020vae,
	title={VAE-SNE: a deep generative model for simultaneous dimensionality reduction and clustering},
	author={Graving, Jacob M and Couzin, Iain D},
	journal={BioRxiv},
	year={2020},
	publisher={Cold Spring Harbor Laboratory}
}

License

Released under a Apache 2.0 License. See LICENSE for details.