Working Paper: CEPR ID: DP14024
Authors: Lilia Maliar; Serguei Maliar; Pablo Winant
Abstract: Artificial intelligence (AI) has impressive applications in many fields (speech recognition, computer vision, etc.). This paper demonstrates that AI can be also used to analyze complex and high-dimensional dynamic economic models. We show how to convert three fundamental objects of economic dynamics -- lifetime reward, Bellman equation and Euler equation -- into objective functions suitable for deep learning (DL). We introduce all-in-one integration technique that makes the stochastic gradient unbiased for the constructed objective functions. We show how to use neural networks to deal with multicollinearity and perform model reduction in Krusell and Smith's (1998) model in which decision functions depend on thousands of state variables -- we literally feed distributions into neural networks! In our examples, the DL method was reliable, accurate and linearly scalable. Our ubiquitous Python code, built with Dolo and Google TensorFlow platforms, is designed to accommodate a variety of models and applications.
Keywords: artificial intelligence; machine learning; deep learning; neural network; stochastic gradient; dynamic models; dynamic programming; Bellman equation; Euler equation; value function
JEL Codes: No JEL codes provided
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| AI techniques (C45) | ability to analyze complex economic models (E17) |
| conversion of economic dynamics objects (lifetime reward, Bellman equation, Euler equation) (C61) | application of DL techniques (C45) |
| stochastic gradient descent (C69) | efficiency of model solutions (C52) |
| all-in-one integration method (F15) | improved computational outcomes (C63) |
| neural networks (C45) | improvements in model performance (C52) |
| DL method (Y60) | computational efficiencies (C63) |