Phenomenological Recession: From Gaussian to Bayesian Probability

Abstract

If recession is a random event, then it might be governed by probabilistic laws instead determinism, because unexpected confluence of variables showing notable changes in their behavior in time. This paper, proposes the idea that recession might be perceived as a kind of transition from a Gaussian profile to one dictated by Bayesian probabilities. Under this approach, recession exhibits its phenomenological character as observed at the chain of events and the sharpness of variables previous to its main manifestation. This would be seen as a temporal evolving whose interrelation of involved variables might be conditional among each other, triggering a fast increasing in the probability of having contraction of economical activity. The Mitchell criteria inside the framework of Machine Learning to test theoretical proposal were used. Simulations are presented and discussed.