(Universidad Internacional de Andalucía, 2018) Caki, Mamed; Caki, Mamed
Trabajo de Máster en Economía, Finanzas y Computación. Director: Dr. José Manuel Bravo Caro. Economic conditions over long time periods can be distinguished by regimes. Regime identification has been object of numerous investigations in economics and financial modeling for years. Recently, new machine learning technics such as decision trees, support vector machines and neural networks, among others, followed by alternative datasets and cheap computational processing power became available, allowing for alternative ways to model complex economic relationships. In the present work, we develop a supervised machine learning classifier using Random Forest technic to identify economic regimes using the S&P 500 stock market index series.
