Forecasting Recessions With Scikit-Learn

It is no secret that everybody wants to predict recessions. Many economists and finance firms have attempted this with limited success, but by and large there are several well known leading indicators for recessions in the US economy. However, when presented to the general public these indicators are typically taken alone, and are not framed in a way that can give probability statements associated with an upcoming recession. In this project, I have taken several of those economic indicators and built a classification model to generate probabilistic statements. Here, the actual classification ('recession' or'no recession') is not as important as the probability of a recession, since this probability will be used to determine a basic portfolio scheme which I will describe later on.