Ekonomi & Bisnis editorial
Open AccessOA2023
Deep Learning in Economics: A Systematic and Critical Review
This systematic review examines the integration of deep learning into economics, surveying commonly used models and their applications across various subfields. It provides critical insights into implementation strategies, research gaps, and future challenges, highlighting the growing role of deep learning in intelligent economic decision-making.
Yuanhang Zheng; Zeshui Xu; Anran Xiaoยท Artificial Intelligence Reviewยท 2023ยท DOI 10.1007/s10462-022-10272-8
The core problem
The methodology of economics has evolved from qualitative to quantitative approaches, and from small data samples to vast datasets. Deep learning models, with their superior ability to learn inherent laws and representative levels, are increasingly assisting in intelligent decision-making in economics. This paper systematically investigates deep learning in economics, presenting statistical results of relevant research, a survey of frequently-used models, and their applications. It also offers critical reviews on models, applications, implementation strategies, research gaps, and future challenges. The authors note that deep learning has been widely applied in various economic subfields, including financial economics, macroeconomics and monetary economics, agricultural and natural resource economics, industrial organization, urban, rural, regional, real estate and transportation economics, health, education and welfare, business administration, and microeconomics. The review aims to provide a comprehensive overview and critical assessment of the field.
Innovation
The authors conducted a systematic literature review of deep learning applications in economics. They surveyed frequently-used deep learning models in economics, such as feedforward neural networks, convolutional neural networks (CNNs), recurrent neural networks (RNNs), long short-term memory (LSTM) networks, and their variants. The review covers applications across multiple subfields of economics. The methodology includes a statistical analysis of relevant researches, possibly examining publication trends, model usage frequencies, and application areas. The paper also provides critical reviews from multiple perspectives: models and applications, why and how to implement deep learning in economics, research gaps, and future challenges. The systematic approach ensures a comprehensive coverage of the literature, though specific details on search strategies, inclusion criteria, or databases used are not provided in the abstract.
Introduction
The methodology of economics has evolved from qualitative to quantitative approaches, and from small data samples to vast datasets. Deep learning models, with their superior ability to learn inherent laws and representative levels, are increasingly assisting in intelligent decision-making in economics. This paper systematically investigates deep learning in economics, presenting statistical results of relevant research, a survey of frequently-used models, and their applications. It also offers critical reviews on models, applications, implementation strategies, research gaps, and future challenges. The authors note that deep learning has been widely applied in various economic subfields, including financial economics, macroeconomics and monetary economics, agricultural and natural resource economics, industrial organization, urban, rural, regional, real estate and transportation economics, health, education and welfare, business administration, and microeconomics. The review aims to provide a comprehensive overview and critical assessment of the field.
Why it matters
The paper provides critical reviews of deep learning in economics, addressing several key aspects. First, it critically examines the models and applications, discussing their strengths and limitations. Second, it explores why and how to implement deep learning in economics, likely discussing the advantages of deep learning over traditional econometric methods, such as handling large datasets, capturing nonlinear relationships, and improving predictive accuracy. However, challenges such as interpretability, data quality, and overfitting are also considered. Third, the review identifies research gaps, such as the need for more theoretical foundations, better model interpretability, and integration with economic theory. Finally, future challenges are discussed, including ethical considerations, computational requirements, and the need for interdisciplinary collaboration. The authors conclude that deep learning will play an increasingly important role in economic decision-making, but further research is needed to address these challenges.
Who should read this
Pembaca Research Digest dan praktisi rumpun ini.
Opening member contentโฆ