Working for the Ecole polytechnique fédérale de Lausanne (EPFL) means being part of a prestigious school that consistently ranks among the top 20 universities worldwide. As a public university dedicated to improving the world around us, we have three missions: training, research and technology transfer. We boast one of the most dynamic university campuses in Europe and employ more than 6,000 people. Our employees perform high-value-added work in teaching and research and in the school’s administrative and technical services. Between our main campus in Lausanne and our satellite campuses in Geneva, Neuchâtel, Fribourg and Sion, our workforce is composed of more than one hundred different professions.

At EPFL, we foster a culture of respect and inclusion in the workplace. We promote a healthy work-life balance through flexible working hours and on-campus daycare and sports facilities. Our employees also benefit from belonging to a diverse community of 16,000 people – including over 10,000 students and 3,500 researchers – from 120 different countries.

Post Doctoral Researcher in the Information Processing Group of the School of Computer and Communication Sciences at EPFL

Your mission :

A two-years post-doctoral position is available in the framework of a collaborative project between Nicolas Macris (EPF Lausanne), Florent Krzakala and Guilhem Semerjian (ENS Paris) on Phase Diagrams and Algorithms for Inference and Learning - PAIL:



The fields of high dimensional statistics, modern inference and machine learning have witnessed transformative changes in the last few years. In particular, recent algorithmic advances on deep networks have convincingly shown that it is possible to learn non-trivial features in data in an unsupervised way. However, these exciting advances are almost all empirical. Although the new approaches find their roots in neural network systems; there is very little theoretical basis or understanding for their recent successes at discovering and correctly identifying features or good representations of data.



A growing body of work suggests that we can understand and locate fundamental information theoretic and algorithmic barriers by thinking of them as phase transitions (e.g. in the sense of physics). The last decade has seen a convergence of successful methods, principles and mathematical theories in the disciplines of coding theory, random combinatorial optimization, signal processing and statistical mechanics of disordered systems. These connections are currently witnessing an impressive renaissance.



Whether we consider detecting rare and subtle effects in dispersed massive databases, identifying widely scattered communities in large networks, reaching the information-theoretic limits of sensing, coding, and data transmission, removing noise from images and signals, or predicting high dimensional systems, the most exciting and promising recent theoretical developments share intriguing common features.



The aim of the present interdisciplinary project is to benefit from the recent developments in fields as diverse as theoretical statistical physics coding, signal processing, probability and random combinatorial optimization to uncover principles hidden behind the success of the new approaches in modern inference problems.




Main duties and responsibilities include :

• The post-doctoral researcher will be a member of the Lab for Communication Theory in EPFL. He will collaborate with the members and PhD students of the lab (and beyond) within the scope of the PAIL project. The collaboration will involve short term visits in Paris with F. Krzakala and G. Semerjian. The opportunity is given to work with initiative on a wide range of modern problems in inference and learning. We will appreciate a team-player in research.

• In terms of teaching we expect some participation in student project supervision. Opportunities to teach an advanced subject or run a reading group exist according to taste.


• There are no administrative duties.

Your profile :

• The project is interdisciplinary by nature and our team members have diverse backgrounds. Background and PhD degree in one of the following disciplines is welcome: applied mathematics, probability, theoretical physics, spin-glass physics, information theory, signal processing, theoretical computer science or machine learning.

• Some knowledge in more than one of the above disciplines is a plus. A strong taste for theory required. We also expect reasonable programming expertise to implement algorithmic ideas.

We offer :

• The post-doctoral researcher will have the opportunity to work and interact in scientifically strong and stimulating environments within the School of Computer and Communications sciences in EPFL as well visit the partners of the project in Ecole Normale Supérieure in Paris.

 The researcher will also benefit from our international contacts in other institutions.

• We offer first class working conditions with state-of-the-art equipment. Sufficient travel funds are available within the PAIL project in order to participate to international conferences, for collaborative visits in other institutions, winter and summer schools.

• The current project funds the post-doctoral position for two years according to EPFL salary scale (initial contract is for one-year renewable an extra year upon agreement of both parties).

Start date :

To be agreed upon but earlier the better before 1 April 2019.
Term of employment :

Fixed-term (CDD)
Work rate :

100%
Duration :

CDD one year renewable one extra yearContact :

For additional information, please contact Dr. Nicolas Macris, MER by e-mail nicolas.macris@epfl.ch



Please send your application (cover letter, detailed resume and publication list)



apply online