Timezone: »

 
Poster
Fast $(1+\varepsilon)$-Approximation Algorithms for Binary Matrix Factorization
Ameya Velingker · Maximilian Vötsch · David Woodruff · Samson Zhou

Tue Jul 25 05:00 PM -- 06:30 PM (PDT) @ Exhibit Hall 1 #432
in Poster Session 2 »
We introduce efficient $(1+\varepsilon)$-approximation algorithms for the binary matrix factorization (BMF) problem, where the inputs are a matrix $\mathbf{A}\in\{0,1\}^{n\times d}$, a rank parameter $k>0$, as well as an accuracy parameter $\varepsilon>0$, and the goal is to approximate $\mathbf{A}$ as a product of low-rank factors $\mathbf{U}\in\{0,1\}^{n\times k}$ and $\mathbf{V}\in\{0,1\}^{k\times d}$. Equivalently, we want to find $\mathbf{U}$ and $\mathbf{V}$ that minimize the Frobenius loss $\|\mathbf{U}\mathbf{V} - \mathbf{A}\|_F^2$. Before this work, the state-of-the-art for this problem was the approximation algorithm of Kumar et. al. [ICML 2019], which achieves a $C$-approximation for some constant $C\ge 576$. We give the first $(1+\varepsilon)$-approximation algorithm using running time singly exponential in $k$, where $k$ is typically a small integer. Our techniques generalize to other common variants of the BMF problem, admitting bicriteria $(1+\varepsilon)$-approximation algorithms for $L_p$ loss functions and the setting where matrix operations are performed in $\mathbb{F}_2$. Our approach can be implemented in standard big data models, such as the streaming or distributed models.
[ PDF]  [ Poster] 

Author Information

Ameya Velingker (Google Research)
Maximilian Vötsch (University of Vienna)
David Woodruff (Carnegie Mellon University)
Samson Zhou (UC Berkeley)

More from the Same Authors