Near Parameter Free Ant Colony Optimisation Export

@incollection{citeulike:5291355, abstract = {Ant colony optimisation, like all other meta-heuristic search processes, requires a set of parameters in order to solve combinatorial problems. These parameters are often tuned by hand by the researcher to a set that seems to work well for the problem under study or a standard set from the literature. However, it is possible to integrate a parameter search process within the running of the meta-heuristic without incurring an undue computational overhead. In this paper, ant colony optimisation is used to evolve suitable parameter values (using its own optimisation processes) while it is solving combinatorial problems. The results reveal for the travelling salesman and quadratic assignment problems that the use of the augmented solver generally performs well against one that uses a standard set of parameter values. This is attributed to the fact that parameter values suitable for the particular problem instance can be automatically derived and varied throughout the search process.}, author = {Randall, Marcus}, citeulike-article-id = {5291355}, citeulike-linkout-0 = {http://www.springerlink.com/content/dwrvkbnq1juk37bw}, journal = {Ant Colony, Optimization and Swarm Intelligence}, keywords = {2004, ant-system, optimization}, pages = {374--381}, posted-at = {2009-07-28 15:17:31}, priority = {2}, title = {Near Parameter Free Ant Colony Optimisation}, url = {http://www.springerlink.com/content/dwrvkbnq1juk37bw}, year = {2004} }

TY - CHAP ID - citeulike:5291355 L3 - citeulike-article-id:5291355 N2 - Ant colony optimisation, like all other meta-heuristic search processes, requires a set of parameters in order to solve combinatorial problems. These parameters are often tuned by hand by the researcher to a set that seems to work well for the problem under study or a standard set from the literature. However, it is possible to integrate a parameter search process within the running of the meta-heuristic without incurring an undue computational overhead. In this paper, ant colony optimisation is used to evolve suitable parameter values (using its own optimisation processes) while it is solving combinatorial problems. The results reveal for the travelling salesman and quadratic assignment problems that the use of the augmented solver generally performs well against one that uses a standard set of parameter values. This is attributed to the fact that parameter values suitable for the particular problem instance can be automatically derived and varied throughout the search process. JF - Ant Colony, Optimization and Swarm Intelligence EP - 381 TI - Near Parameter Free Ant Colony Optimisation SP - 374 KW - 2004 KW - ant-system KW - optimization AU - Randall, Marcus PY - 2004/// UR - http://www.springerlink.com/content/dwrvkbnq1juk37bw ER -