Overview

Description

CALENDAR DESCRIPTION:

The first part of the course will focus on dynamic optimization problems, with an emphasis on dynamic programming. Applications may include growth, business cycles, monetary and fiscal policy, and optimal contracts. The second part of the course will focus on models of learning and bounded rationality. Genetic and stochastic approximation algorithms will be studied. Applications may include the stability of rational expectations equilibria, the evolution of institutions and social conventions, and models of robust control and Knightian uncertainty.

COURSE DETAILS:

This graduate course offers a dive into Computational Economics, emphasizing the use of numerical simulations and both parametric and nonparametric structural modelling to unpack complex economic systems. Students will explore parametric and nonparametric structural approaches to market competition and international trade. Cutting-edge nonparametric methods will be explored, focusing on the entropic latent variable integration for demand modelling and nonparametric finite mixture models that allow rich heterogeneity of consumers. The class also integrates Machine Learning and Deep Learning, teaching students to deploy neural networks and clustering for economic analysis.

The expected outcome of this class is to produce a short paper using computational economics tools applied to any economic problem.

Topics: (Not Exhaustive)
1. Numerical and Simulation Methods
2. Parametric Structural Modelling
   2.1 Models of Differentiated Competition
   2.2 Gravity Models of Trade
3. Nonparametric Structural Modelling
   3.1. Nonparametric Modelling of Demand using Entropic Latent Variable Integration Via Simulation
   3.2. Nonparametric Modelling of Heterogeneous Decision Makers using Finite Mixture Techniques
4. Machine Learning and Deep Learning
   4.1 Deep Feedforward Networks
   4.2 Clustering Methods in Economics

Grading

Materials

REQUIRED READING:

REQUIRED READING NOTES: