Küçük açık ekonomi modellerinin genetik algoritma ile eğitilmiş sinirsel ağ ile yakınsanması

This thesis work presents a direct numerical solution methodology to approximatethe small open economy models with debt elastic interest rate premiumand with convex portfolio adjustment cost, both studied by Stephanie Schmitt-Grohe and Martin Uribe(2003). This recent method is compared with the rstorderapproximation to the policy function from the aspect of second momentsof endogenous variables and their impulse responses. The proposed methodology,namely genetic algorithm-neural network (GA-NN), parameterizes thepolicy function with a feed-forward neural network that is trained by a geneticalgorithm. Thus, unlike the rst-order approximation, GA-NN does notrequire the continuity and the existence of derivatives of objective and policyfunctions. Importantly, since genetic algorithm is an evolutionary algorithmthat enables global search over the feasible set, it provides a robust result inany solution space. Also GA-NN method gives not only the moments of themodel but also the optimal path.