Matrix Chain Scheduling Problem: Difference between revisions
Jump to navigation
Jump to search
No edit summary |
No edit summary |
||
Line 5: | Line 5: | ||
== Related Problems == | == Related Problems == | ||
Subproblem: [[Approximate MCSP]] | Subproblem: [[Approximate MCSP]] | ||
Related: [[Approximate MCOP]] | Related: [[Matrix Chain Ordering Problem]], [[Approximate MCOP]] | ||
== Parameters == | == Parameters == |
Latest revision as of 08:18, 10 April 2023
Description
The Matrix Chain Scheduling Problem (or MCSP) is an optimization problem� where the goal is to find the product sequence for evaluating a chain of matrix products and the processor schedule for the sequence such that the evaluation time is minimized on a parallel system.
Related Problems
Subproblem: Approximate MCSP
Related: Matrix Chain Ordering Problem, Approximate MCOP
Parameters
$P$: number of processors
$n$: number of matrices
Table of Algorithms
Name | Year | Time | Space | Approximation Factor | Model | Reference |
---|---|---|---|---|---|---|
Czumaj | 1993 | $O(log^{3} n)$ | $O(n^{2})$? | Exact | Parallel | Time |
References/Citation
https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.54.9426&rep=rep1&type=pdf
https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.56.222&rep=rep1&type=pdf