Bayesian models for forecasting in the supply chain


The global economy relies heavily on the consumption of goods of all kinds: food, textiles, electronics, equipment, vehicles, etc. The supply chain is the mechanism that accounts for the physical flows of these consumed goods. The supply chain consists of logistics (warehouses, transport vehicles, packaging), finance (billing and payment), and IT (stock status, manufacturing, ordering and shipping). Its efficient functioning has a direct impact on the performance of the economy, on the environment (transport, storage, destruction of useless stock) and on the working conditions of the employees. In this thesis, the objective is to predict the evolution of the outputs in the supply chain, i.e. quantities in locations of storage or sales. The research work of this thesis can be divided into two parts: (1) mathematical modeling of the supply chain, and (2) inference methods for prediction. First, we will model the problem in a probabilistic and generic way, taking into account all possible interactions between the variables of interest. In the second part of this thesis, we will develop Bayesian methods for the probabilistic prediction at all stages of the chain. We will propose novel efficient and accurate algorithms, including but not only sequential learning, in order to avoid the need to reprocess all past data every time new data is available.


One fully funded PhD position is available in Lille from September/October 2018. The thesis will take place jointly at the engineering school IMT Lille Douai and at a fast-growing company focused on machine learning and data science. Earlier start date can be also considered. Lille is a vibrant, young and dynamic city. Lille lies in the heart of the triangle that links three of Europe’s main metropoles: London (80 min), Paris (60 min), and Brussels (35 min).


The students will be supervised by:

– Víctor Elvira: http://pagesperso.telecom-lille.fr/elvira/
– François Septier: http://pagesperso.telecom-lille.fr/septier/


We are looking for a motivated and talented student with:
– background in machine learning, signal processing, statistics or applied mathematics
– strong mathematical skills

-experience in programming, preferably in Matlab and/or Python.


If you have any question and/or want to apply, please contact:

– Víctor Elvira: victor.elvira@imt-lille-douai.fr
– François Septier: francois.septier@imt-lille-douai.fr