1 code implementation • 11 Sep 2022 • Matthew Fahrbach, Thomas Fu, Mehrdad Ghadiri
By extending our approach to block-design matrices where one block is a Kronecker product, we also achieve subquadratic-time algorithms for (1) Kronecker ridge regression and (2) updating the factor matrices of a Tucker decomposition in ALS, which is not a pure Kronecker regression problem, thereby improving the running time of all steps of Tucker ALS.
no code implementations • 22 Jul 2021 • Matthew Fahrbach, Mehrdad Ghadiri, Thomas Fu
Low-rank tensor decomposition generalizes low-rank matrix approximation and is a powerful technique for discovering low-dimensional structure in high-dimensional data.
no code implementations • NeurIPS 2019 • Lin Chen, Hossein Esfandiari, Thomas Fu, Vahab S. Mirrokni
In this paper, we aim to develop LSH schemes for distance functions that measure the distance between two probability distributions, particularly for f-divergences as well as a generalization to capture mutual information loss.
no code implementations • 30 Apr 2019 • Mohammadhossein Bateni, Lin Chen, Hossein Esfandiari, Thomas Fu, Vahab S. Mirrokni, Afshin Rostamizadeh
To achieve this, we introduce a novel re-parametrization of the mutual information objective, which we prove is submodular, and design a data structure to query the submodular function in amortized $O(\log n )$ time (where $n$ is the input vocabulary size).