[L2S/CentraleSupélec] PhD position in Machine Learning - See details in the offers.
Signals and Statistics group @L2S
PhD offers in Machine Learning
In a fruitful research environment, we offer PhD positions in the area of machine learning. In partnership with several companies, one develops methodologies adapted to high-dimensional and heterogeneous data. More precisely, we are working on advanced clustering techniques, recommendation procedures, robust statistical learning, models fitting.
One PhD topic concerns recommendations and predictions. More precisely, one aims at building tailored recommendation systems building both on historical data and on metadata. This can be applied to:
- Automatically create smart collections: user oriented and perfumer-oriented selection of formulas/ingredients.
- Tests/briefs prediction thanks to an accurate probability of win.
- Recommendation via reformulation that will be closely related to the first research topic.
To achieve these objectives, the PhD candidate will study and work with modern tools such as similarity methods, communities detections techniques, neural networks and restricted Boltzmann machines networks or matrix factorization.
Some classical references are:
- Robert Bell, Yehuda Koren, Chris Volinsky, “Matrix factorization techniques for recommender systems”, pp. 30-37, vol. 42, Computer, 2009 - computer.org
- Trevor Hastie, Robert Tibshirani, Jerome Friedman, “The Elements of Statistical Learning: Data Mining, Inference, and Prediction”, Second Edition February, Springer, 2009.
- Graham W. Taylor A Geoffrey E. Hinton, “Factored conditional restricted Boltzmann Machines for modeling motion style”, Proceedings of the 26th Annual International Conference on Machine Learning, 978-1-60558-516-1, Montreal, Quebec, Canada.
- Asja Fischer, Christian Igel, Training restricted Boltzmann machines: An introduction, In Pattern Recognition, Volume 47, Issue 1, 2014, pp. 25-39, ISSN 0031-3203.
- Mansinghka, Vikash; Shafto, Patrick; Jonas, Eric; Petschulat, Cap; Gasner, Max; Tenenbaum, Joshua B., “CrossCat: A Fully Bayesian Nonparametric Method for Analyzing Heterogeneous, High Dimensional Data”, Journal of Machine Learning Research 17 (2016) pp. 1-49.
If you are interested, please contact us: