Remy Garnier

I am a PhD student in machine learning and statistics at Analyse Geométrie Modelisation (CREST) in Cergy-Poontoise University.
My PhD Thesis is a collaboration between a company, CDiscount, based in Bordeaux, and the laboratory. My main research topics consists in the application of time series methods for the prediction of long-term demand for different products. More specifically, i am interested in method to perform multi-time series forecasting, both from a theoritical or a practical point of vue. Current method of interest :

  • High-dimensionnal Auto-regressive vectorial models
  • Agregation of predictors
  • Recurrent Neural Networks
My PhD thesis is done under the supervsision of Paul Doukhan and Bruno Goutorbe


Publications

Published

2019 P. Alquier , K. Bertin, P. Doukhan, R. Garnier, High dimensional VAR with low rank transition in Statistics and Computing. Pdf.

2020 R. Garnier Modelisation of competition between time series submitted in Annals of Operation Research Preprint .


Talks

2019 R. Garnier , A. Belletoile A multi-series framework for demand forecasts in E-commerce at APIA 2019. Preprint


Education

Université Paris-Saclay

Master 'Mathématique de l'Aléatoire'
Statistiques - Machine Learning
2017 - 2018

ENS Paris-Saclay

Parisian Master in Computer Science Research
Computer Science-Algorithmics
2014 - 2018