简介：Thecross-fertilizationbetweenartificialintelligenceandcomputationalfinancehasresultedinsomeofthemostactiveresearchareasinfinancialengineering.Onedirectionistheapplicationofmachinelearningtechniquestopricingfinancialproducts,whichiscertainlyoneofthemostcomplexissuesinfinance.Intheliterature,whentheinterestrate,themeanrateofreturnandthevolatilityoftheunderlyingassetfollowgeneralstochasticprocesses,theexactsolutionisusuallynotavailable.Inthispaper,weshallillustratehowgeneticalgorithms(GAs),asanumericalapproach,canbepotentiallyhelpfulindealingwithpricing.Inparticular,wetesttheperformanceofbasicgeneticalgorithmsbyusingittothedeterminationofpricesofAsianoptions,whoseexactsolutionsisknownfromBlack-Scholesoptionpricingtheory.Thesolutionsfoundbybasicgeneticalgorithmsarecomparedwiththeexactsolution,andtheperformanceofGAsisewluatedaccordingly.Basedontheseewluations,somelimitationsofGAsinoptionpricingareexaminedandpossibleextensionstofutureworksarealsoproposed.