PhD candidate in efficient deep learning for weakly labelled data @ University of Amsterdam (the Netherlands)

[Rianne van den Berg]

Vacancy: PhD candidate in efficient deep learning for weakly labelled data
--------------------------------------------------------------------------
The Informatics Institute (IvI) of the Faculty of Science invites applications for a PhD candidate in Efficient Deep Learning for weakly labelled data and High-Tech Systems, for a period of four years. The position is part of the Dutch STW Efficient Deep Learning program. You will be based in the Amsterdam Machine Learning Lab (AMLab) led by Prof. Max Welling within the Informatics Institute. The research will be supervised by Prof. Max Welling, and Dr Rianne van den Berg. You will also cooperate with Prof. H. Corporaal in the Electronic Systems group at Eindhoven University of Technology (TU/E). Furthermore, you will spend part of your time at Qualcomm Research Netherlands.

Application closing date: 1 October 2018
Preferred starting date: between October 2018 and February 2019
Duration: 4 years


Project description
---------------------
The Efficient Deep Learning (EDL) program combines the fields of machine learning and computing. Both disciplines are strongly represented in the Netherlands and are now connected by seven Dutch academic institutes and more than 35 other (industrial) partners in- and outside the Netherlands. The EDL program contains seven use-case driven EDL research projects: P1) DL as a service, P2) Reconstruction, matching and recognition, P3) Video analyses and surveillance, P4) High tech systems and materials, P5) Human and animal health, P6) Mobile robotics, and P7) DL platforms. The common goal for all seven EDL projects is to significantly improve the applicability of DL among others by creating data efficient training, and improving computational efficiency, both for training and inference.

As PhD candidate you will be part of the EDL project P4 for deep learning for high-tech systems and materials (HTSM). Partners in this project are the Dutch universities TU/e, UvA and VU, as well as Thermo Fischer, Qualcomm, the Netherlands eScience center, ASTRON and SURFsara. For this position, the main collaborating organizations will be the University of Amsterdam, TU/E and Qualcomm.
You will work as part of the Amsterdam Machine Learning Lab on deep learning methods for HTSM problems with large complex data streams. These data streams are often characterized by large volumes of unlabelled data, with only a small portion of labelled/annotated data points, and extremely unbalanced clusters. You will focus on active learning for efficient labelling strategies, as well as semi-supervised learning. Active learning for weakly labelled data can be aided by calibrated uncertainty estimates in model predictions. Improving uncertainty estimates for deep learning models will thus be a large part of the project.

Requirements
-------------------
We are looking for candidates whose skill set matches with (a large part of) the following profile:
- a Master's degree in artificial intelligence, computer science, mathematics, statistics, or closely related area;
- affinity with machine learning and deep learning;
- excellent programming skills (in C, C++, Python);
- excellent mathematical skills (especially in probability theory and statistics, calculus, and linear algebra);
- affinity with signal processing;
- experience with High-Performance Computing (parallel programming, networking, GPUs);
- a team player who enjoys a multicultural and interdisciplinary environment in which academic-industrial collaboration is central;
- commitment and a cooperative attitude;
- strong communication, presentation and writing skills and excellent command of English.

Further information
-------------------
For further information, including instructions on submitting an application,
please see the official job advertisement at:

http://www.uva.nl/en/content/vacancies/2018/09/18-527--phd-candidate-in-efficient-deep-learning-for-weakly-labelled-data.html?a

Informal inquiries can be made by email to Rianne van den Berg (r.vand...@uva.nl).


-------------------------------------------------------------
Rianne van den Berg | Postdoctoral Researcher
Amsterdam Machine Learning Lab (AMLab)
Informatics Institute | University of Amsterdam
-------------------------------------------------------------
Science Park 904 | 1098 XH Amsterdam
+31 (0)20.525.8426
https://riannevdberg.github.io
-------------------------------------------------------------