Genetic Algorithm and Direct Search Toolbox™

Parallel Computing with gamultiobj

Parallel computing with gamultiobj works almost exactly the same as with ga. For detailed information, see Parallel Computing with the Genetic Algorithm.

The difference between parallel computing with gamultiobj and ga has to do with the hybrid function. gamultiobj allows only one hybrid function, fgoalattain. This function is optionally launched after gamultiobj finishes its run. Each individual in the calculated Pareto frontier, i.e., the final population found by gamultiobj, becomes the starting point for an optimization using fgoalattain. These optimizations are done in parallel. The number of processors performing these optimizations is the smaller of the number of individuals and the size of your matlabpool.

For fgoalattain to run in parallel, its options must be set correctly:

fgoalopts = optimset('UseParallel','always')
gaoptions = gaoptimset('HybridFcn',{@fgoalattain,fgoalopts});

Run gamultiobj with gaoptions, and fgoalattain runs in parallel. For more information about setting the hybrid function, see Hybrid Function Options.

gamultiobj calls fgoalattain using a parfor loop, so fgoalattain does not estimate gradients in parallel when used as a hybrid function with gamultiobj. This is because a parfor loop estimates gradients in parallel, and inner iterations in a nested parfor loops do not run in parallel.

Using gamultiobj

References