PUBLISHED PAPERS #09.03
| Tamerlan Ibrahimzade. Offering New Insights for Behavioral Economics Via Machine Learning |
|---|
| Abstract. Machine learning methods have been widely utilized across fields, delivering enhanced computational performance and uncovering hidden data patterns. In economics, particularly behavioral economics, their application depends on factors such as dataset nature, relationship complexity, interpretability, computational resources, and research questions. Deep learning, specifically deep neural networks, has revolutionized artificial intelligence. Based on the Turing-Church understanding of computation, these interactive methods surpass traditional algorithmic approaches like "divide and conquer" or dynamic programming, mimicking cognitive processes and enabling a deeper exploration of bounded rationality. Additionally, we examine recent applications of deep reinforcement learning (DRL) in economics. As a subset of AI, DRL effectively tackles complex problems through interactive learning, offering powerful tools for economic analysis. Balancing model complexity and interpretability is critical. This study focuses on personal consumption expenditures from a public dataset, employing Linear Regression and Decision Trees in Python for comparison. |
| Keywords: machine learning, deep learning, behavioral economics, decision trees, linear regression, personal consumption expenditures, deep neural networks, deep reinforcement learning |
 Download PDF |
| DOI: https://doi.org/10.30546/MaCoSEP2025.1050 |