MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

13
MRPGA An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter 古古古

Transcript of MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

Page 1: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm

Reporter:古乃卉

Page 2: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

Outline

Abstract Introduction Related Work Architecture MRPGA Implementation Experiments Conclusion

Page 3: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

Abstract

MapReduce Map and Reduce

Genetic Algorithm Iteration

MRPGA Extension of MapReduce for Parallelizing

Genetic Algorithm

Page 4: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

Introduction

Problems of Parallelized Genetic Algorithm Communication, synchronization,

heterogeneity and frequent failures Why MapReduce?

Provides a parallel design pattern for simplify application developments

How to work? Add a phase for global selection at the end of

every iteration of PGAs and a coordinator

Page 5: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

Related Work

PGAs Distributed, coarse grained and fine grained MPI : not flexible enough for handling

heterogeneity and failures MapReduce

Phoenix, Hadoop and MRPSO

Page 6: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

Architecture

Page 7: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

MRPGA

Map, Reduce and Reduce

Key : index of the individual Value : the individual Allows each of the reduce tasks to collect

dependent input without fetching data from a remote machine

Page 8: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

Key : individual Value : just number

MRPGA(cont.)

Page 9: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

MRPGA(cont.)

Select the global Optimum individual

Reproduction, mutationand submission ofoffspring to thescheduler of MRPGA, and collection optimumindividual

Page 10: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

Implementation

Page 11: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

Experiments

MRPGA runtime system with Aneka An enterprise Grid consisting of 33 nodes Pentium 4 processor 1GB of memory 160GB IDE disk 1 Gbps Ethernet Windows XP

Page 12: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

300 individuals 100 generations

Simulated cost Avg. evaluation 10 sec. Standard deviation 0.2

500 individuals 10 times

Experiments(cont.)

MOAE

MOAE+MRPGA

Page 13: MRPGA : An Extension of MapReduce for Parallelizing Genetic Algorithm Reporter :古乃卉.

Conclusion

This extension makes PGAs can benefit from the MapReduce model on handling heterogeneity and failures