The Unreasonable Effectiveness of Random Orthogonal Embeddings

@inproceedings{Choromanski2017TheUE,
  title={The Unreasonable Effectiveness of Random Orthogonal Embeddings},
  author={Krzysztof Choromanski and Mark Rowland and Adrian Weller},
  year={2017}
}
We present a general class of embeddings based on structured random matrices with orthogonal rows which can be applied in many machine learning applications including dimensionality reduction, kernel approximation and locality-sensitive hashing. We show that this class yields improvements over previous stateof-the-art methods either in computational efficiency (while providing similar accuracy) or in accuracy, or both. In particular, we propose the Orthogonal Johnson-Lindenstrauss Transform… CONTINUE READING
2 Citations
18 References
Similar Papers

References

Publications referenced by this paper.

Similar Papers

Loading similar papers…