Fast spectral algorithms from sum-of-squares proofs: tensor decomposition and planted sparse vectors
STOC '16: Symposium on Theory of Computing Cambridge MA USA June, 2016, pp. 178-191, 2016.
We consider two problems that arise in machine learning applications: the problem of recovering a planted sparse vector in a random linear subspace and the problem of decomposing a random low-rank overcomplete 3-tensor. For both problems, the best known guarantees are based on the sum-of-squares method. We develop new algorithms inspired ...More
PPT (Upload PPT)