Nonlinear Versus Linear Learning Devices: A Procedural Perspective

摘要: We provide a discussion of bounded rationality learning behind traditional mechanisms, i.e., Recursive Ordinary Least Squares and Bayesian Learning . These mechanisms lack for many reasons behavioral interpretation and, following the Simon criticism, they appear to be ’substantively rational‘. In this paper, analyzing Cagan model, we explore two which more plausible from point view somehow ’procedurally rational‘: Mean linear models Back Propagation Artificial Neural Networks The algorithms look minimum variance error forecasting by means steepest descent gradient procedure. analysis model shows an interesting result: non-convergence Rational Expectations Equilibrium is not due restriction devices; also may fail converge model.