Mixture models, robustness, and sum of squares proofs

Samuel B. Hopkins, Jerry Li

2018 · 111 citations · 51 references

DOIFull text

Open access

Concepts

Abstract

We use the Sum of Squares method to develop new efficient algorithms for learning well-separated mixtures of Gaussians and robust mean estimation, both in high dimensions, that substantially improve upon the statistical guarantees achieved by previous efficient algorithms. Our contributions are:

References

51