Anthony Coache  —  Homepage

Anthony Coache 

Master student in Statistics

I am currently a Master Student in Mathematics, Concentration in Statistics at Université du Québec à Montréal (UQÀM) advised by Prof. François Watier and Prof. René Ferland. I graduated with Honours from the B. Sc. in Mathematics, Concentration in Statistics from UQÀM in 2017. I have a strong background in mathematics and programming.

I am a teaching assistant for statistics and mathematics courses and private tutor for several mathematics courses.

I co-organized the Statistics Student Summit in Montreal that was held on March 15, 2019.

Research interests

I am interested in multidisciplinary problems at the intersection of statistics and computer science. More specifically, my research interests are in stochastic modelling, optimization, machine learning, applied statistics and computer science.



  • Binette, O. & Coache, A. The Significance of the Adjusted R Squared. (Bio)Statistics Research Day, Montréal. September 21, 2018. [Poster] [Blog]

  • Coache, A. & Larose, F. “Do schools kill creativity?” Well, they help analyze popularity! Annual Meeting of the SSC, Montréal. June 4, 2018. [Poster]

  • Ferland, R., Froda, S. & Coache, A. Comparison of surveillance flu data across regions. Annual Meeting of the SSC, Winnipeg. June 12, 2017. [Poster]



  • Stochastic Algorithms for Solving a Multi-period Quantile-Based Portfolio Optimization Problem. Annual Meeting of the SSC, Calgary. May 27, 2019. [Slides] [Beamer template]

    • In financial asset allocation, while many risk measures are proposed, care must be taken to ensure that their properties reflect an observed investor behavior. Furthermore, the investor should also be able to adjust his strategy according to market fluctuations during the investment period. Thus we focus our analysis on a class of multiperiod portfolio optimization problems in which the investor wants to minimize a quantile-based function of the terminal wealth and where the stock prices follow a binomial tree model. We explore various stochastic approaches in finding an optimal or near-optimal strategy and compare their efficiency through simulation studies.

  • Non-Parametric Estimation of the Quantile Function. Probability and Statistics Student Seminar - UQAM. July 13, 2017.

    • This talk focuses on the non-parametric estimation of the quantile function. The quantile function has several statistical applications; it is used in Monte-Carlo methods among others. We explain two L-estimators of this function and we compare their performance against the traditional non-parametric estimator.


Cellphone: (514)-946-9661

Université du Québec à Montréal
Pavillon Président-Kennedy