Econometrics, Social Interactions and Social Networks, Time Series
I am currently a Ph.D candidate in economics at Université Laval. My thesis is supervised by two awesome professors of the UL economic department, Vincent BOUCHER (my main Adviser) and Bernard FORTIN (my co-Adviser). I also often work with Arnaud Dufays and one of my thesis chapter is a joint work with him. Previously, I got a B.SC and a M.SC in Statistics. Please, download my Curriculum Vitae for more details.
Regarding my thesis, I mainly work on Econometric of Social Networks and Time series topics.
In one of my thesis chapter, we propose a new method for estimating peer effect when the full network is not observed or partially observed. We use the true network distribution that can be estimated with any network formation model. With this distribution, we find a way to simulate the unobserved network and valid instruments to estimate the linear in mean peer effects model. We also provide a Bayesian framework in which, the true network distribution acts as the network prior distribution for the unobserved network.
In another chapter, we propose a variables selection method to detect relevant changes in AR and ARX models parameters. When structural breaks occur in the data, we determine which parameters abruptly change. As soon as there are several breaks or explanatory variables, the number of models to consider becomes large and find the parameters that change at each break is computationally cumbersome. We then propose a penalty function to explore the space of these models. We eventually associate posterior probabilities to the most likely models.
I am currently working (alone) on my last thesis essay. In this chapter, I propose a model of social network with unlimited counting data.
Please, visit my research page to learn more about my work.
I am also passionate about programming. Please visit my "Media and Code" page to learn more.