The closing date for this job submission has passed.

Job Description

This postdoctoral position takes part in a TOTAL research project aiming to develop new predictive model for special fuel manufacturing. Each product may have both a molecular signa- ture, obtained by chromatographic analysis, and physicochemical properties. The main objective of this post-doctoral position is to establish links between both quantities by means of supervised learning techniques. This objective can come in two axes : (i) selection of an informative sample of observations among massive set of observations and (ii) selection of an informative small subset of variables among a large data set.
Preliminary studies have been done in order to test the reliability of sparse supervised learning (sparse regression, sparse SVM,...). The objective of this post-doctoral position is thus dedicated to develop both new models and new algorithms in order to improve the existing predictive model. Moreover, in order to select observations of interest, we will focus on recent developments on ac- tive learning using submodular informative criteria. The theoretical results could be published in international conferences and journals.

Job Information

email redacted
Related URL
Laboratoire de Physique ENSL
Topic Category
Lyon, France
Closing Date
March 1, 2016