Economic policy making using neural networks

Neural networks are introduced for economic policy making. A desirable property of a policy instrument is its speedy impact on the economic system. Monetary policy is generally recognized as the one, relative to the fiscal policy, which is easier to implement and its impact is faster to realize. An application of the simultaneous perturbation stochastic approximation-based neural networks to monetary policy formulation for the US economy shows that its policy is faster to respond to sudden changes in the dynamic of the system than a traditional linear feedback policy