Computational Economics and Machine Learning for Solving Dynamic Economic Models Using Matlab and Python
The course will introduce all the basic tools for numerical analysis of equilibria in dynamic economic models including conventional computational-economic methods and recent machine learning from the field of data science.
By the end of the course the participants will have acquired detailed knowledge of and hands-on experience in:
- Key concepts and notions of economic dynamics (Bellman equation, Euler equation, Markov decision process, state and control variables, recursive equilibrium, non-stationary models, etc.).
- Conventional computational methods for analyzing dynamic economic models, including perturbation, projection and stochastic simulation approaches.
- Methods of computational economics for ameliorating the curse of dimensionality including sparse grids, ergodic set methods, non-product integration, quasi-Monte Carlo methods, endogenous grid method, parallelization, etc.
- Machine learning tools from the fields of supervised, unsupervised and reinforcement learning, including linear and logistic regression, cluster analysis, dimensionality reduction, decision trees, principal component analysis, stochastic optimization, etc.
- Applications of neural networks and deep learning methods for analyzing high dimensional economic models including heterogeneous-agent new Keynesian model with incomplete markets, large scale models of central banking.
- Matlab and python computer codes for analyzing dynamic economic models including deep learning libraries such as TensorFlow.
The course combines a theoretical presentation of tools of computational economics and machine learning with practical hands-on experience in designing and implementing computer codes for solving dynamic economic models. The participants will be using both Matlab and python languages including the state-of-the-art deep learning libraries such as TensorFlow. Our code will be ubiquitous and portable to other applications the participants might be interested in.
Instructor:
Serguei Maliar, currently a professor at the Leavey School of Business, Santa Clara University, arrived from Stanford University where he was a Visiting Associate Professor in the Department of Economics and a Visiting Fellow for the Hoover Institution. He has taught at the University of Alicante, University of Pompeu Fabra and University of Chicago (ICE program). He also served as the Director of Graduate Studies in Economics at the University of Alicante.
Professor Maliar's specialization is in macroeconomics, economic theory, game theory, numerical methods, economies in transition, and economic growth and development. His scholarship has appeared in numerous top journals such as Econometrica, Quantitative Economics, Review of Economic Dynamics, Journal of Economic Dynamic and Control, and Journal of Business and Economic Statistics; he contributed a chapter to Handbook of Computational Economics; and is currently advising Canadian Central Bank on the model for the optimal monetary choice. He is an associate editor of Journal of Economic Dynamic and Control and a recipient of NSF grant for 2016-2019 academic years.
Professor Maliar earned his B.S in Physics and Applied Mathematics from Moscow Institute of Physics and Technology; his M.A. in Economics from Central European University (Czech Republic); his Ph.D. in Applied Mathematics from Zaporozhye State University (Ukraine); and his Ph.D. in Economics from the University of Pompeu Fabra (Spain).
Registration: EcoMod School Registration 2025