Resumen

MORALES-CASTROI, Arturo; RAMIREZ-REYES, Eliseo  y  SANABRIA-LANDAZABAL, Néstor Juan. MEXICAN STOCK EXCHANGE PERFORMANCE AFTER THE FINANCIAL CRISIS OF 2008: APPLICATION OF DATA MINING. Dimens.empres. [online]. 2020, vol.18, n.1, pp.28-38. ISSN 1692-8563.  https://doi.org/10.15665/dem.v18i(1).2246.

Machine learning prediction models are explored to analyze the performance of the Mexican Stock Exchange (PQI) after the 2008 crisis. These models have shown good forecasting capabilities both for multivariable and univariable approaches given their non-parametric features. Financial variables representative of the Exchange was selected. The models were evaluated with Mean Absolute Percent Error (MAPE) metric and compared to Linear Regression (LR) and Neural Nets (NN) models. The results show that the models had similar performance according to the error percentages they presented to those of LR and NN; in some cases, with better performance.

Palabras clave : stock exchange performance; financial crisis; data mining.

        · resumen en Español | Portugués     · texto en Inglés     · Inglés ( pdf )