|
Tobias Oechtering
News
Master thesis projects: I am looking for qualified master students from KTH that would like to do a master thesis project with me in one of the following three areas: (i) privacy mechanisms for data disclosure with provable privacy guarantees (either theory or applied),
(ii) study on memorisation in learning to meet right-to-be-forgotten as in GDPR, (iii) design of trustworthy digital twins for energy management or predictive maintenance.
All projects are part of larger research activities. Please contact me if you are interested and have the right background.
New video tutorial: Privacy-preserving data analysis: How can we learn from data while protecting sensitive information? Jointly with Sara Saeidian, I explore this question in “An Introduction to Privacy-Preserving Data Analysis with Applications in AI.” Almost five hours of videos take you from the foundations to pointwise maximal leakage, a privacy concept recently developed in our research. Whether you are new to privacy or want to understand the ideas behind our work, you are welcome to take a look!
EECS Faculty Board: I have been elected to the EECS Faculty Board for January 2027–December 2029. Thank you for your trust—I look forward to contributing to the school’s work. (Election results)
Dig-IT Lab summer school: Jointly with Qianwen Xu, I organised the first online Dig-IT Lab summer school on Aug 13–21, 2026. The school brought together 216 participants from 55 countries to explore digitalisation of the built environment, combining lectures, hands-on sessions, and interactive group work. Read more in the Dig-IT Lab news article.
Empirical PML estimation: In our recent paper on empirical estimation of PML, we show how the PML privacy risk can empirically estimated from data. Ths was previously considered to be a weakness of PML. The paper will be presented at IEEE CSF 2026, a preprint is available on
arXiv:2509.22428.
End of KTH Digitalisation platforms: In the digitalisation domain, collaborations across schools and with society has reached a fantastic level at KTH so that it has been decided to discontinue the research platforms by mid 2024. Therewith ends also my Digitalisation platform director assignment after five exciting years.
Workshop Trustworthy Learning: On Oct 16, 2023, Digital Futures DataLEASH and EXTREMUM projects organise a workshop on “Trustworthy Learning.” More information and registration can be found here link.
Workshop Digitalisation in Energy research: On Sept 14, 2023, KTH Digitalisation and Energy research platforms organize workshop “Data-driven methods with energy applications.” More information and registration can be found here link.
Pointwise Maximal Leakage (PML): Our foundational paper introducing PML has been accepted in July 2023 to be published in IEEE Transactions on Information Theory, PML is a promising novel robust privacy measure that allows provable quantitative privacy risk assessment, preprint available arXiv:2205.04935.
Workshop Digitalisation in Transport research: On Sept 20, 2022, KTH Digitalisation and Transport research platforms organize workshop “Autonomous transportation - Challenge in digitalisation and solution to sustainability.” More information and registration can be found here link.
Workshop Security for Industry: On May 23, 2021, KTH Digitalisation and Industrial Transformation research platforms organize workshop “Security for Industry - How to avoid a Cyber Attack.” More information and registration can be found here link.
VR project grant: The Swedish Research council has granted our exciting project proposal “Fundamental bounds for Distributed Decision-making,” with collaborator Mael Le Treust, CNRS researcher as co-PI, starting 2021 where we will re-visit the Witsenhausen problem with a fresh approach. Our initial results building on our initial works such as ITW 2019 paper already reveal interesting insights.
Workshop AI for Life Science: On Oct 14, 2020, KTH Digitalisation and Life Science research platforms organize workshop “AI for Life Science.” More information and registration can be found here link.
VR project grant: In Nov 2016, I got informed that my project proposal on distributed source coding with security and privacy submitted to the Swedish Research Council (VR) has been granted. Soon, I am going to look for a new PhD student to work with me on this exciting project.
This homepage has been initially set up in autumn 2016, is irregularly maintained by Tobias Oechtering, and contains only personal information and views.
|