we are looking for a post-doc in the area of large scale sequential learning, to work with Dr. Jérémie Mary and myself.
The position is open in the SequeL research group, located in Lille, France (see http://sequel.lille.inria.fr/).
The duration of the contract is 1 year, starting no later than 28 Dec. 2012. Due to administrative procedures, applying is urgent.
The candidate should be eager to contribute to the state of the art of ML, either on fundamental issues, or the design of new and efficient algorithms along with their implementation to deal with today large amounts of data (say O(10^10), or more).
He/she has a passion for bringing sound, mathematically grounded ideas, into working algorithms to face this amount of data.
The topic of the work is voluntarily left rather open: it can be related to matrix and tensor factorization, gradient methods, sparsity issues, graphs, …The sequential aspect is important: the data the algorithm are supposed to process are evolving along time (imagine the graph of the web, information flowing from sensors, …). On top of that, decision may be taken based on what has been learned, turning the whole thing into a sequential decision problem. (This decision making aspect is not absolutly necessary to deal with during the post-doc: it is left as an open direction).
Please note that regarding implementations, Matlab and that sort of software are not welcome. We look forward real software developped in C/C++. Prototyping may be done in R.
For more details about the offer, and the application procedure, please visit
If you wish to apply, please email me and provide a detailed CV.