Efficient parallelizations of a competitive learning algorithm for text retrieval on the MasPar

Author

Syu, Inien ; Lang, S.D. ; Hua, Kien A.

Author_Institution

Dept. of Comput. Sci., Central Florida Univ., Orlando, FL, USA

fYear

1995

fDate

6-9 Feb 1995

Firstpage

4

Lastpage

11

Abstract

In this paper, we present parallel implementations of a connectionist model for text retrieval on the MasPar MP-1, an SIMD machine with up to 16 K processors. The connectionist model was originally developed on a SUN SparcStation 1+ for a sequential implementation. In our parallel implementations, we consider three strategies for mapping the network onto the MasPar: one-to-one, many-to-one, and one-to-many, depending on the ratio of the network size to the number of processors, in order to reduce the computation time. We also consider load balancing among processors for further improvement in performance. Our experimental results demonstrate noticeable speedups in our parallel implementations

Keywords

information retrieval; neural nets; parallel processing; unsupervised learning; MasPar MP-1; SIMD machine; competitive learning algorithm; connectionist model; load balancing; parallel implementations; parallelizations; performance evaluation; text retrieval; Computer networks; Computer science; Concurrent computing; Databases; Information retrieval; Load management; Lungs; Neural networks; Partitioning algorithms; Sun;

fLanguage

English

Publisher

ieee

Conference_Titel

Frontiers of Massively Parallel Computation, 1995. Proceedings. Frontiers '95., Fifth Symposium on the

Conference_Location

McLean, VA

Print_ISBN

0-8186-6965-9

Type

conf

DOI

10.1109/FMPC.1995.380470

Filename

380470