*Job opening*: 1 post-doc position in Machine Learning and Natural Language
Processing at Instituto Superior Tecnico, University of Lisbon, Portugal
*Project*: DeepSPIN ("Deep Structured Prediction in Natural Language
Processing"), ERC Starting Grant (2018-2023)
*Application deadline*: July 31, 2020
*Job description*:
There is one open position for a postdoctoral researcher in the scope of
the ERC project DeepSPIN ("Deep Structured Prediction in Natural Language
Processing"). The overarching goal of DeepSPIN is to bring together deep
learning and structured prediction to solve challenging tasks in natural
language processing (e.g., machine translation, visiolinguistic tasks,
syntax/semantic parsing). Three research directions are:
- Better planning mechanisms for structural reasoning. This includes
efficient use of large contexts and long-term memories, better handling of
uncertainty, and research on calibration and confidence estimation methods.
- Induction of interpretable latent structure, exploiting synergies with
probabilistic inference and sparse modeling. This includes sparse and
structured attention mechanisms, latent variable models, and induction of
latent explanations.
- Weak supervision and few-shot learning to reduce the need for labeled
data. This includes discrete, semi-supervised latent variable models and
modular networks to learn disentangled representations.
See https://deep-spin.github.io/ for more information about DeepSPIN, our
team, and some of our recent publications.
The specific roles assigned to the project members will take into account
their research interests and their background.
The post-doc is a 2-year position, with the possibility of extension up to
3 years. The salary range is 2500-3000 EUR (free of tax), depending on the
candidate's experience. This post-doc position can potentially evolve into
a “ELLIS post-doc” and include a visit to an exchange institution in the
ELLIS network (https://ellis.eu/phd-postdoc).
If interested in this position, please send an email with your CV to
[email protected]. Feel free to contact me for any further
information.
*Working environment*:
The host institution (Instituto de Telecomunicações) is a leading R&D
institution (having hosted 60+ FP7/H2020 projects) located in the campus of
Instituto Superior Técnico (IST) at the University of Lisbon, where I am an
Associate Professor. IST is the top engineering school in Portugal, with
11,000+ students and 700+ faculty members, currently hosting 4 ERC grants.
The host institution is part of the Lisbon Unit for Machine Learning and
Intelligent Systems (LUMLIS) and part of the ELLIS network (
https://ellis.eu/units). Collaborations with other ELLIS members are
encouraged.
Part of the work may be done in collaboration with Unbabel, a start-up
translation company with unique assets (a crowd of 10,000+ human
translators, and post-edited translations in 50+ language pairs).
Lisbon is one of the oldest cities in Western Europe, being the major
political, economic and cultural centre of Portugal. It enjoys a
Mediterranean climate, having one of the warmest winters of any metropolis
in Europe and a long summer season lasting from June to September. Portugal
has an outstanding coastline and is home to some of the most stunning
beaches in Europe, being a popular spot for surfers and other water sport
enthusiasts. Video:
https://www.youtube.com/watch?list=UUKrO-9IoZLYqCC7a-rfuNYQ&v=R2c6r5aiMuU.
Lisbon is also home to the Lisbon Machine Learning School (LxMLS,
http://lxmls.it.pt), which every year brings together 200+ students and the
world leading researchers in machine learning and natural language
processing.
More information: https://andre-martins.github.io/pages/jobs.html.
_______________________________________________
uai mailing list
[email protected]
https://it.engineering.oregonstate.edu/mailman/listinfo/uai
