SPECULATIVE ASYNCHRONOUS SUB-POPULATION EVOLUTIONARY COMPUTING

• US 20130006901A1
• Filed: 06/30/2011
• Published: 01/03/2013
• Est. Priority Date: 06/30/2011
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Cantin, Jason F.
• Granted Patent: US 9,563,844 B2
• Time in Patent Office: Days
• US Class Current: 706/13
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/126 Evolutionary algorithms, e.g. genetic algorithms or genetic programming

Abstract

A tool computes fitness values for a first generation of a first sub-population of a plurality of sub-populations. A population of candidate solutions for an optimization problem was previously divided into the plurality of sub-populations. The population of candidate solutions was created for an iterative computing process in accordance with an evolutionary algorithm to identify a most fit candidate solution for the optimization problem. The tool determines a speculative ranking of the first generation of the first sub-population prior to the fitness values being computed for all candidate solutions in the first generation of the first sub-population. The tool generates a next generation of the first sub-population based, at least in part, on the speculative ranking prior to completion of computation of the fitness values for the first generation of the first sub-population.

Claims

1-8. -8. (canceled)

9. A computer program product for speculative evolutionary computing, the computer program product comprising:

a computer readable storage medium having computer usable program code embodied therewith, the computer usable program code comprising a computer usable program code configured to;

compute fitness values for a first generation of a first sub-population of a plurality of sub-populations, wherein a population of candidate solutions for an optimization problem was previously divided into the plurality of sub-populations, wherein the population of candidate solutions was created for an iterative computing process in accordance with an evolutionary algorithm to identify a most fit candidate solution for the optimization problem;

determine a speculative ranking of the first generation of the first sub-population prior to the fitness values being computed for all candidate solutions in the first generation of the first sub-population; and

generate a next generation of the first sub-population based, at least in part, on the speculative ranking prior to completion of computation of the fitness values for the first generation of the first sub-population.

10. The computer program product of claim 9, wherein the computer usable program code is further configured to:

speculate that the fitness values of the first generation of the first sub-population will not prevent a cataclysm for the first sub-population prior to completing computation of the fitness values of the first generation of the first sub-population,

wherein the computer usable program code configured to generate the next generation of the first sub-population based, at least in part, on the speculative ranking prior to completion of computation of the fitness values for the first generation of the first sub-population comprises the computer usable program code configured to select a set of one or more survivor candidate solutions from the first generation of the first sub-population based, at least, in part, on the speculative ranking.

11. The computer program product of claim 10, wherein the computer usable program code configured to speculate that the fitness values of the first generation of the first sub-population will not prevent the cataclysm for the first sub-population prior to completing computation of the fitness values of the first generation of the first sub-population comprises the computer usable program code configured to speculate that the fitness values of the first generation of the first sub-population will not prevent the cataclysm after computing fitness values for a predefined number of the candidate solutions of the first generation of the first sub-population.

12. The computer program product of claim 9, wherein the computer usable program code configured to generate the next generation of the first sub-population based, at least in part, on the speculative ranking comprises the computer usable program code configured to insert a set of one or more migrant candidate solutions into the next generation of the first sub-population based, at least in part, on the speculative ranking of the first generation.

13. The computer program product of claim 9, wherein the computer usable program code is further configured to select a set of one or more candidate solutions from the first generation of the first sub-population for migration based, at least in part, on the speculative ranking.

14. A computer program product for speculative evolutionary computing, the computer program product comprising:

a computer readable storage medium having computer usable program code embodied therewith, the computer usable program code comprising a computer usable program code configured to;

receive fitness values for a first generation of a first sub-population of a plurality of sub-populations, wherein a population of candidate solutions for an optimization problem was previously divided into the plurality of sub-populations, wherein the population of candidate solutions was created for an iterative computing process in accordance with an evolutionary algorithm to identify a most fit candidate solution for the optimization problem;

determine whether the first generation of the first sub-population does not satisfy a termination criterion for the iterative computing processing;

determine whether the first generation corresponds to a later iteration of the iterative computing process than a second generation of a second sub-population of the plurality of sub-populations;

determine whether a difference between the first generation and the second generation does not exceed a termination speculation threshold; and

generate a third generation of the first sub-population responsive to a determination that the difference between the first generation and the second generation does not exceed the termination speculation threshold, wherein the computer usable program code is configured to generate the third generation of the first sub-population based, at least in part, on the fitness values.

15. The computer program product of claim 14, wherein the computer usable program code is configured to generate the third generation of the first sub-population in accordance with one or more evolutionary computing techniques implemented in accordance with the evolutionary algorithm.

16. The computer program product of claim 15, wherein the evolutionary computing techniques comprise at least one of crossover, mutation, and reproduction.

17. The computer program product of claim 14, wherein the computer usable program code is further configured to divide the population of candidate solutions for the optimization problem into the plurality of sub-populations of candidate solutions and to assign the plurality of sub-populations to computing resources.

18. The computer program product of claim 14, wherein the computer usable program code configured to determining whether the first generation corresponds to a later iteration of the iterative computing process than the second generation of the second sub-population of the plurality of sub-populations comprises the computer usable program code configured to:

access generation tracking data based on a first computing resource identifier and a second computing resource identifier,

wherein the generation tracking data indicates an iteration of the iterative computing process for each of a plurality of computing resources,

wherein the first computing resource identifier identifies a first of the plurality of computing resources and a second computing resource identifier identifies a second of the plurality of computing resources,

wherein the first computing resource is associated with the first sub-population and the second computing resource is associated with the second sub-population; and

compare a first iteration indicated for the first computing resource and a second iteration indicated for the second computing resource identifier.

19. The computer program product of claim 14, wherein the computer usable program code is configured to:

determine that the third generation of the first sub-population should receive a migrant candidate solution from a neighboring one of the plurality of sub-populations;

determine that the migrant candidate solution from the neighboring one of the plurality of sub-populations is not available;

create a speculative migrant candidate solution; and

insert the speculative migrant candidate solution into the third generation of the first sub-population of candidate solutions.

20. The computer program product of claim 19, wherein the computer usable program code is configured to determine that a condition for speculatively creating a migrant candidate solution is satisfied before creating the speculative migrant candidate solution.

21. The computer program product of claim 19, wherein the computer usable program code is further configured to tag the speculative migrant candidate solution to indicate speculative.

22. An apparatus comprising:

a processor;

a network interface coupled with the processor; and

a speculative evolutionary computing unit, the speculative evolutionary computing unit operable to;

receive fitness values for a first generation of a first sub-population of a plurality of sub-populations, wherein a population of candidate solutions for an optimization problem was previously divided into the plurality of sub-populations, wherein the population of candidate solutions was created for an iterative computing process in accordance with an evolutionary algorithm to identify a most fit candidate solution for the optimization problem;

determine whether the first generation of the first sub-population does not satisfy a termination criterion for the iterative computing processing;

determine whether the first generation corresponds to a later iteration of the iterative computing process than a second generation of a second sub-population of the plurality of sub-populations;

determine whether a difference between the first generation and the second generation does not exceed a termination speculation threshold; and

generate a third generation of the first sub-population responsive to a determination that the difference between the first generation and the second generation does not exceed the termination speculation threshold, wherein the computer usable program code is configured to generate the third generation of the first sub-population based, at least in part, on the fitness values.

23. The apparatus of claim 22, wherein the speculative evolutionary computing unit is further operable to:

determine that the third generation of the first sub-population should receive a migrant candidate solution from a neighboring one of the plurality of sub-populations;

determine that the migrant candidate solution from the neighboring one of the plurality of sub-populations is not available;

create a speculative migrant candidate solution; and

insert the speculative migrant candidate solution into the third generation of the first sub-population of candidate solutions.

24. The apparatus of claim 22 further comprising a computer readable storage medium having computer usable program code stored therein, wherein the computer usable program code embodies the speculative evolutionary computing unit.