Postdoc position in Safe Machine Learning – Aalborg University, Denmark

    Postdoc position in Safe Machine Learning - Aalborg University, Denmark

    Postdoc position is a part of Swift Project gathering doctoral and postdoctoral researchers around the subject of machine learning, multi-objective optimization, safety, security and privacy preserving algorithms, at the Technical Faculty of IT and Design, Department of Electronic Systems, Aalborg University, Denmark


    Deadline to Apply

    June 20, 2021 (23: 59 GMT +1)


    Overview

    PositionPostdoc position
    No. of Position(s)1
    Research AreaSafe Machine Learning
    SalaryAccording to standard norms
    WorkplaceTechnical Faculty of IT and Design
    Department of Electronic Systems
    Aalborg University
    Denmark
    Contract Period2 Years (with the possibility of extension)
    Starting dateAug 01, 2021

    Qualifications

    PhD Degree in related fields

    Responsibilities/ Job Description

    Machine learning forges its way into many spheres of human activity from education, medical and legal supporting systems to autonomous vehicles. It is capable of classifying objects, trends, and patterns in a huge amount of data. Nonetheless, the control and monitoring solutions for critical infrastructures stay reluctant to this new trend due to high safety standards. To this end, we propose to merge stochastic safety with reinforcement learning. The learning intrinsically makes use of trials and exploration of new situations. Nonetheless, these new unknown situations may become unsafe points with no return.

    Tasks
    We will develop the theory of safety focusing on mathematical characterizations and numerical approximation methods for the associated safety function and safety measure. Specifically, we will identify the set of initial states from which the probability of reaching an unsafe region (before reaching the desired target) is small enough.

    We will consider model-driven optimal control based on dynamic programming. We will introduce stochastic barrier certificates in dynamic programming. This work will pave the way for solutions in data-driven optimization methods. We will use these results for safe dynamic programming to equip Reinforcement Learning with a barrier certificate.

    How to Apply?

    The applications are only to be submitted online by using the “Apply online” button above

    Documents Required

    • A motivated text wherein the reasons for applying, qualifications in relation to the position, and intentions and visions for the position are stated.
    • A current curriculum vitae.
    • Copies of relevant diplomas (Master of Science and PhD). On request you could be asked for an official English translation.
    • Scientific qualifications. A complete list of publications must be attached with an indication of the works the applicant wishes to be considered. You may attach up to 5 publications.
    • Dissemination qualifications, including participation on committees or boards, participation in organisations and the like.
    • Additional qualifications in relation to the position. References/recommendations.
    • Personal data.

    About Research project

    Swift Project
    The position is a part of Swift Project gathering doctoral and postdoctoral researchers around the subject of machine learning, multi-objective optimization, safety, security and privacy preserving algorithms.

    About the Department

    The Department of Electronic Systems employs more than 200 staff, of which about 40% are international, and about 90 enrolled PhD students. The Department hosts 620 students. The Department’s excellent research infrastructures and facilities accentuate its global position in teaching and research. Department research centres around communication, antennas, control systems, signal processing, cyber security and acoustics. The Department plays an active role in translating discoveries and results in practical applications with industrial partners and through IPR. Within its subject field, the Department provides teaching for several BSc and MSc programmes using a problem- and project-based learning model.

    Note

    Employment is in accordance with the Ministerial Order on the Appointment of Academic Staff at Universities (the Appointment Order) and the Ministry of Finance’s current Job Structure for Academic Staff at Universities. Employment and salary are in accordance with the collective agreement for state-employed academics.   

    Inquiries

    Professor Rafal Wisniewski
    Department of Electronic Systems
    Email: raf@es.aau.dk

    Official advertisement

    POSTDOC IN SAFE MACHINE LEARNING

    Post expires at 8:59am on Monday June 21st, 2021 (GMT+9)