Future-proof Masters thesis on AI & Statistical Modelling
Combine Control Systems AB / Matematikerjobb / Göteborg
Observera att sista ansökningsdag har passerat.
Visa alla matematikerjobb i Göteborg,
Mölndal,
Partille,
Kungälv,
Lerum eller i
hela Sverige Visa alla jobb hos Combine Control Systems AB i Göteborg,
Linköping,
Lund eller i
hela Sverige We work with control systems and analysis, model based development and software development. Our assignments contain challenges at the highest technical level and in our engineers you'll find analytical minds, inventive souls and a can-do attitude.
What drives us is a passion for technology and a genuine commitment to people. This is manifested in our open-minded atmosphere, informal culture, and that we prioritize skills above background.
We are located in Gothenburg (HQ), Lund, Linköping.
www.combine.se Are you interested on a future-proof Master's thesis on AI and/or Statistical modelling? Do you have a reliable mathematics foundation? Are you proficient at implementing algorithms in Java/C and Python? Are you a resilient individual who knows how to prioritize his time and energy? Graph theory does not scare you (and in fact, you think is pretty cool)?
Then you are set for success! Also, we might have a thesis or two to tickle your fancies. The joint goal of the thesis is to study Partially Observable Markov Decision Processes with Information Rewards (a.k.a. POMDP-IR). Now, what's that? Well, is a mathematical framework similar to Markov Decision Processes (MDP) in which a robot/agent makes decisions to reach a goal, in spite of not having complete knowledge of the state of the environment. In addition to this, interaction with the environment is possible in order to gather information, and the robot/agent receives a reward given the state of the world and the decision taken. States of the world are modelled with distribution functions.
Who will you work with?
As you can already see, there are two key elements here: The decision making process, and the statistical modelling of the state. Each is a problem of its own, and the kind of problems offered by each task may require a different skillset. Which is why we decided to split this study into two topics.
One thesis will focus on studying the mathematics and implementation of POMDP given the previously described domain. This thesis will be directed by Jennifer Renoux & Pedro U. Lima.
Jennifer Renoux is a post-doctoral fellow in the Machine Perception and Interaction lab at AASS Örebro University. She has worked on multi-agent active information gathering, planning and reasoning under uncertainty and multiple E-health and defense projects.
Pedro U. Lima is a Professor at IST, Universidade de Lisboa, and a researcher of the Institute for Systems and Robotics where he coordinates the Intelligent Robots and Systems group. Pedro is interested in discrete event models of robot tasks and planning under uncertainty, with applications to networked robot systems. He is very active in robot competitions, and a Trustee of the RoboCup Federation (2003-2012, 2016-), General Chair of RoboCup2004, and member of the Advisory Board of the Mohamed Bin Zayed International Robotics Challenge (MBZIRC), whose first edition took place in the United Arab Emirates in March 2017.
The second thesis will study the mathematics and implementation of the statistic models that the POMDP will use. This thesis will be directed by Lia Silva-Lopez.
Lia Silva-Lopez is the Manager for Data Science Solutions of Combine Control Systems. She has worked on statistical modelling for automotive applications, and is a co-editor of Statsletters.com. She is interested in applying mathematics, statistics and algorithms to make everyday life better.
What are your skills?
For the AI topic:
+ Computer Science background.
+ Understanding or strong interest on Planning.
+ Understanding of basic probability theory.
+ Competent on C-like languages/Java AND Python.
+ Can write tidy code.
For the Statistical Models topic:
+ Understanding and strong interest in the mathematics of probability.
+ Has implemented/used exiting statistical models in Python (on your own or in class).
+ Competent in Python.
+ Can write tidy code.
Nice to have on both topics:
* Worked with ROS or find it absolutely delightful to work with.
* Understanding of Markov decision processes.
Mindset:
+ Open minded. You want to learn!
+ You like bouncing back thoughts and working with other people. Especially an international team with renowned researchers.
+ You are polite and actively make sure that your argumentation style cannot easily be traced to any in this list.
+ You are self-motivated.
+ You are willing to write the thesis and work in English.
Meet the Robots!
Depending on your specific set of competences, we might apply the framework on different domains. We could agree on working with a customer from Combine Control Systems, or we might work with Robots. Meet the robots!
If you have the foundations and the will to work on either topic, we encourage you to apply! Good knowledge on both AI and Statistical modelling will allow you to benefit from this thesis the most. So go tell all your like-minded pals about this opening!
For more information, feel free to contact Lia Silva-Lopez at
lia.silva@combine.se or 031 797 1015.
The project takes place in the Göteborg office of Combine Control Systems, next to Linnégatan in Järntorget 4, an artsy district of Gothenburg lined with cafés, restaurants, pubs and the most interesting variety of shops and theaters. It is easily reachable by trams, buses and ferry.
Combine Control Systems prides itself on the quality of its people and its relaxed atmosphere. We are a leading supplier of premium engineering services within data analysis, control systems and software development. Our goal is to help our customers bring their products to the market on time, at a predictable cost and with high quality. We offer varied challenges at the bleeding edge of technology, an inspiring working environment and enthusiastic colleagues. We have offices in Gothenburg, Lund and Linköping. Come check us out!
Varaktighet, arbetstid
• -
Publiceringsdatum2017-12-04ErsättningMasters thesis
Så ansöker duSista dag att ansöka är 2017-12-14
Klicka på denna länk för att göra din ansökanFöretagCombine Control Systems AB
AdressCombine Control Systems AB
Järntorget 4
41304 Gothenburg
KontorsadressJärntorget 4
Jobbnummer 3834880
Observera att sista ansökningsdag har passerat.