Asymmetric Random Projections

06/22/2019
by   Nick Ryder, et al.
0

Random projections (RP) are a popular tool for reducing dimensionality while preserving local geometry. In many applications the data set to be projected is given to us in advance, yet the current RP techniques do not make use of information about the data. In this paper, we provide a computationally light way to extract statistics from the data that allows designing a data dependent RP with superior performance compared to data-oblivious RP. We tackle scenarios such as matrix multiplication and linear regression/classification in which we wish to estimate inner products between pairs of vectors from two possibly different sources. Our technique takes advantage of the difference between the sources and is provably superior to oblivious RPs. Additionally, we provide extensive experiments comparing RPs with our approach showing significant performance lifts in fast matrix multiplication, regression and classification problems.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
09/23/2021

Sparse Regression Faster than d^ω

The current complexity of regression is nearly linear in the complexity ...
research
10/18/2022

Faster Matrix Multiplication via Asymmetric Hashing

Fast matrix multiplication is one of the most fundamental problems in al...
research
06/01/2023

Some New Non-Commutative Matrix Multiplication Algorithms of Size (n,m,6)

For various 2≤ n,m ≤ 6, we propose some new algorithms for multiplying a...
research
11/25/2022

Secure Distributed Gram Matrix Multiplication

The Gram matrix of a matrix A is defined as AA^T (or A^TA). Computing th...
research
01/11/2021

Enumeration Algorithms for Conjunctive Queries with Projection

We investigate the enumeration of query results for an important subset ...

Please sign up or login with your details

Forgot password? Click here to reset