Linear Computation Coding

by   Ralf Müller, et al.

We introduce the new concept of computation coding. Similar to how rate-distortion theory is concerned with the lossy compression of data, computation coding deals with the lossy computation of functions. Particularizing to linear functions, we present an algorithm to reduce the computational cost of multiplying an arbitrary given matrix with an unknown column vector. The algorithm decomposes the given matrix into the product of codebook wiring matrices whose entries are either zero or signed integer powers of two. For a typical implementation of deep neural networks, the proposed algorithm reduces the number of required addition units several times. To achieve the accuracy of 16-bit signed integer arithmetic for 4k-vectors, no multipliers and only 1.5 adders per matrix entry are needed.



page 1

page 2

page 3

page 4


A fast algorithm for computing the Smith normal form with multipliers for a nonsingular integer matrix

A Las Vegas randomized algorithm is given to compute the Smith multiplie...

Efficient Matrix Multiplication: The Sparse Power-of-2 Factorization

We present an algorithm to reduce the computational effort for the multi...

Binary Matrix Factorisation and Completion via Integer Programming

Binary matrix factorisation is an essential tool for identifying discret...

Accuracy Controlled Structure-Preserving H^2-Matrix-Matrix Product in Linear Complexity with Change of Cluster Bases

H^2-matrix constitutes a general mathematical framework for efficient c...

Straggler Mitigation through Unequal Error Protection for Distributed Approximate Matrix Multiplication

Large-scale machine learning and data mining methods routinely distribut...

Computation of the Adjoint Matrix

The best method for computing the adjoint matrix of an order n matrix in...

Characterization of matrices with bounded Graver bases and depth parameters and applications to integer programming

An intensive line of research on fixed parameter tractability of integer...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.