Four Essays on Machine Learning in Microeconomics and Macroeconomics

The role of Machine Learning techniques in Economics has increased rapidly in recent years. This is mainly due to the increasing amount of available data as a consequence of technological progress in collecting and storing them and, most of all, to the diffusion of the internet. As a matter of fact, the internet contains real time information capable of being exploited from an economic perspective. Once those data are stored, to economists is left the task of interpreting them in order to investigate and predict market's evolutions and agents' behavior. This is generally called Nowcasting. For this reason, in the last decade the use of alternative (or non-orthodox) data- sources, models and techniques aimed to study economical phenomena and the cooperation between economic departments and the computer sciences' ones have been encouraged and addressed. An overview of the applications of data and text mining techniques in economics and finance will be given. The role of text mining on predicting the effects of monetary policies will be stressed and further applications of such techniques in the field of forward guidance from central banks will be proposed.