Editing Computational economics (section)

=== Machine learning in computational economics ===
[[Machine learning|Machine learning models]] present a method to resolve vast, complex, unstructured data sets. Various machine learning methods such as the [[kernel method]] and [[random forest]] have been developed and utilized in [[Data mining|data-mining]] and statistical analysis. These models provide superior classification, predictive capabilities, flexibility compared to traditional statistical models, such as that of the [[STAR model|STAR]] method. Other methods, such as causal machine learning and [[Causal model|causal tree]], provide distinct advantages, including inference testing.

There are notable advantages and disadvantages of utilizing machine learning tools in economic research. In economics, a model is selected and analyzed at once. The economic research would select a model based on principle, then test/analyze the model with data, followed by [[Cross-validation (statistics)|cross-validation]] with other models. On the other hand, machine learning models have built in "tuning" effects. As the model conducts empirical analysis, it cross-validates, estimates, and compares various models concurrently. This process may yield more robust estimates than those of the traditional ones.

Traditional economics partially normalize the data based on existing principles, while machine learning presents a more positive/empirical approach to model fitting. Although Machine Learning excels at classification, predication and evaluating goodness of fit, many models lack the capacity for statistical inference, which are of greater interest to economic researchers. Machine learning models' limitations means that economists utilizing machine learning would need to develop strategies for robust, [[Causal inference|statistical causal inference]], a core focus of modern empirical research. For example, economics researchers might hope to identify [[Confounding|confounders]], [[confidence interval]]s, and other parameters that are not well-specified in Machine Learning algorithms.<ref name=":93">{{Citation |title=The Impact of Machine Learning on Economics |date=2019 |url=http://dx.doi.org/10.7208/chicago/9780226613475.003.0021 |work=The Economics of Artificial Intelligence |pages=507–552 |publisher=University of Chicago Press |doi=10.7208/chicago/9780226613475.003.0021 |isbn=9780226613338 |s2cid=67460253 |access-date=2022-05-05|last1=Athey |first1=Susan |url-access=subscription }}</ref>

Machine learning may effectively enable the development of more complicated [[Heterogeneity in economics|heterogeneous]] economic models. Traditionally, heterogeneous models required extensive computational work. Since heterogeneity could be differences in tastes, beliefs, abilities, skills or constraints, optimizing a heterogeneous model is a lot more tedious than the homogeneous approach (representative agent).<ref name=":103">{{Cite book |last=Jesus |first=Browning, Martin Carro |url=http://worldcat.org/oclc/1225293761 |title=Heterogeneity and microeconometrics modelling |date=2006 |publisher=CAM, Centre for Applied Microeconometrics |oclc=1225293761}}</ref> The development of reinforced learning and deep learning may significantly reduce the complexity of heterogeneous analysis, creating models that better reflect agents' behaviors in the economy.<ref name=":113">{{Cite journal |last1=Charpentier |first1=Arthur |last2=Élie |first2=Romuald |last3=Remlinger |first3=Carl |date=2021-04-23 |title=Reinforcement Learning in Economics and Finance |url=https://doi.org/10.1007/s10614-021-10119-4 |journal=Computational Economics |language=en |doi=10.1007/s10614-021-10119-4 |arxiv=2003.10014 |s2cid=214612371 |issn=1572-9974}}</ref>

The adoption and implementation of [[neural network]]s, [[deep learning]] in the field of computational economics may reduce the redundant work of [[Data cleansing|data cleaning]] and data analytics, significantly lowering the time and cost of large scale data analytics and enabling researchers to collect, analyze data on a great scale.<ref name=":73">{{Cite journal |last1=Farrell |first1=Max H. |last2=Liang |first2=Tengyuan |last3=Misra |first3=Sanjog |date=2021 |title=Deep Neural Networks for Estimation and Inference |journal=Econometrica |volume=89 |issue=1 |pages=181–213 |doi=10.3982/ecta16901 |s2cid=203696381 |issn=0012-9682|doi-access=free |arxiv=1809.09953 }}</ref> This would encourage economic researchers to explore new modeling methods. In addition, reduced emphasis on data analysis would enable researchers to focus more on subject matters such as causal inference, confounding variables, and realism of the model. Under the proper guidance, machine learning models may accelerate the process of developing accurate, applicable economics through large scale empirical data analysis and computation.<ref name=":83">{{Cite journal |date=2021-07-27 |title=Deep learning for individual heterogeneity: an automatic inference framework |doi=10.47004/wp.cem.2021.2921 |s2cid=236428783 |doi-access=free }}</ref>