cv
The main elements of my curriculum vitae are provided below. A more complete PDF version is available for download.
Basics
| Name | Sylvain Barde |
| Label | Computational Economist |
| s.barde@kent.ac.uk | |
| Phone | +44 1 227 824 092 |
| Url | https://sylvain-barde.github.io/ |
Work
-
2010.09 - Current -
2010.08 - 2007.09 Chargé d'Etudes (Research Economist)
Observatoire Français des Conjonctures Economiques
Research analyst working on economic geography, innovation diffusion and industrial policy
Education
Awards
- 2001.2002
Chevening Scholar
Foreign, Commonwealth and Development Office, UK
Chevening scholarships are the UK government's global scholarship programme, funded by the Foreign, Commonwealth and Development Office (FCDO), and are awarded to outstanding scholars with leadership potential.
Publications
-
2024.04.23 Bayesian estimation of large-scale simulation models with Gaussian process regression surrogates
Computational Statistics and Data Analysis
This paper develops a likelihood-free bayesian estimation method for large-scale simulation models (in particular Agent-Based Models) that is effective even when the compute budget for simulating the model is limited.
-
2020.02.01 Macroeconomic simulation comparison with a multivariate extension of the Markov information criterion
Journal of Economic Dynamics and Control
This paper develops a multivariate extension of the previous Markov Information Criterion, and shows that it can reliably be used for model comparison/selection in the context of macroeconomic simulation models.
-
2016.11.01 Direct comparison of agent-based models of herding in financial markets
Journal of Economic Dynamics and Control
This paper provides an empirical application of the MIC methodology, comparing the performance of a set of popular agent-bsed models of financial volatility against a benchmark set of ARCH/GARCH processes.
-
2016.10.03 A practical, accurate, information criterion for Nth order Markov processes
Computational Economics
This paper develops an information criterion for Markov processes which can be calculated using only sequences of simulated data from candidate models, enabling accurate comparison of performance across a wide range of model classes.
Skills
| Python | |
| General-purpose programming | |
| numpy/scipy computational applications | |
| torch/Pytorch for machine learning |
| Jupyter | |
| Interactive notebook widgets | |
| RISE interactive slideshows | |
| Data visualisation |
| C++ | |
| ABM simulation models | |
| Cython/Python integration |
| HTML | |
| Web development | |
| Jupyter/RISE integration |
| MATLAB | |
| Computational applications | |
| Teaching |
| Stata | |
| Econometric applications |
| QGIS | |
| GIS/spatial applications |
Languages
| French | |
| Native speaker |
| English | |
| Native speaker |
| Spanish | |
| Fluent |
Interests
| Economics | |
| Economic Geography | |
| Computational Economics | |
| Agent-based models |
| Econometrics | |
| Spatial Econometrics | |
| Bayesian Econometrics | |
| Likelihood-free inference |