Artificial Intelligence From a Behavioral Finance Perspective

Abstract

The non-linear, non-normally distributed, and often dynamic nature of financial data makes the use of predictive analytical techniques widespread. Digital data, which is the most common type of data in finance, has a multidimensional, noisy, and complex data structure that makes data processing and analysis difficult by its very nature. In recent years, the assumption that decision-makers behave rationally in financial decisions has been the subject of “behavioral finance” and empirical studies have proven that many behavioral anomalies affect financial decisions. In order to improve the success of financial forecasting, this chapter proposes to incorporate decision-making behavior into the financial big data structure. Thus, the authors discuss how predictive analytical approaches can successfully forecast not only a noisy and multidimensional data structure but also a multivariate data structure that includes behavioral factors. This study focuses on how AI applications should be used to overcome all of these challenges related to financial research.