DocumentCode :
3034993
Title :
Fast algorithms for matrix multiplication using pseudo-number-theoretic transforms
Author :
Yagle, Andrew E.
Author_Institution :
Dept. of Electr. Eng. & Comput. Sci., Michigan Univ., Ann Arbor, MI, USA
Volume :
2
fYear :
1995
fDate :
9-12 May 1995
Firstpage :
1145
Abstract :
We provide a novel approach to the design of fast algorithms for matrix multiplication. The operation of matrix multiplication is reformulated as a convolution, which is implemented using pseudo-number-theoretic transforms. Writing the convolution as multiplication of polynomials evaluated off the unit circle reduces the number of multiplications without producing any error, since the (integer) elements of the product matrix are known to be bounded. The new algorithms are somewhat analogous to the arbitrary precision approximation (APA) algorithms, but have the following advantages: (1) a simple design procedure is specified for them; (2) they do not suffer from roundoff error; and (3) reasons for their existence is clear. The new algorithms are also non-commutative, so that they may be applied recursively to block matrix multiplication. This work establishes a link between matrix multiplication and fast convolution algorithms and so opens another line of enquiry for the fast matrix multiplication problem. Some numerical examples illustrate the operation of the new proposed algorithms
Keywords :
convolution; matrix multiplication; number theory; polynomials; transforms; arbitrary precision approximation; block matrix multiplication; design procedure; fast convolution algorithms; fast matrix multiplication; noncommutative algorithms; polynomials; product matrix; pseudo-number-theoretic transforms; unit circle; Algorithm design and analysis; Approximation algorithms; Convolution; Joining processes; Polynomials; Roundoff errors; Writing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1995. ICASSP-95., 1995 International Conference on
Conference_Location :
Detroit, MI
ISSN :
1520-6149
Print_ISBN :
0-7803-2431-5
Type :
conf
DOI :
10.1109/ICASSP.1995.480438
Filename :
480438
Link To Document :
بازگشت