Machine Learning in Finance: A Theoretical Review of Its Applications

Abstract

The application of machine learning in finance has been the subject of various studies; however, an analysis of its specific areas of application, its economic contributions, and the limitations surrounding its implementation warrants further exploration. This article offers a theoretical synthesis that connects these various dimensions, providing a better understanding of the possibilities and constraints associated with integrating such technologies into financial activities. It aims to present the key contributions from this rapidly evolving field of research, examine the diversity of its applications, and identify the challenges and future prospects that may shape subsequent research. The study addresses the conceptual foundations of artificial intelligence, machine learning, and deep learning, as well as the application of these techniques to financial operations. It also highlights the opportunities these methods provide for leveraging complex financial data, identifying non-linear relationships, and supporting decision-making processes, while simultaneously acknowledging the primary limitations of their use. Methodologically, this article adopts a descriptive theoretical synthesis approach, based on the analysis and contextualization of scholarly work dedicated to the application of machine learning in finance. This approach emphasizes a conceptual and comparative review of major contributions, with the goal of identifying the areas where these technologies are deployed, their economic benefits, and the limitations inherent in their integration into financial activities.

Classification JEL: O33, G17, C53

Paper type: Theoretical Research