PhD position in Generalization Driven Distributed Continual Machine Learning

Organisation/Company

Uppsala universitet

Department

Uppsala University, Disciplinary Domain of Science and Technology, Faculty of Technology, Department of Electrical Engineering

Research Field

Engineering

Researcher Profile

First Stage Researcher (R1)

Country

Sweden

Application Deadline

31 Mar 2025 - 21:59 (UTC)

Type of Contract

Temporary

Job Status

Full-time

Is the job funded through the EU Research Framework Programme?

Not funded by a EU programme

Is the Job related to staff position within a Research Infrastructure?

No

Electrification and digitalisation are among the largest areas for the future in the conversion to sustainable societies. The Department of Electrical Engineering conducts successful research and education in the areas - renewable energy sources, electric vehicles, industrial IoT, AI, 6G communication and wireless sensor networks as well as research and education within Life Science, smart electronic sensors and medical systems. The Department of Electrical Engineering is an international workplace with around 160 employees that all contribute to important technical energy and health challenges at the Ångström Laboratory.

The position will be at the Division of Signals and Systems, at the Department of Electrical Engineering. Here you will find a friendly work environment and strong research projects. The Division of Signals and Systems collaborates with Swedish companies - public and private - and stakeholders in the different fields of research. We look forward to receiving your application. Join us and build the future with us!

About the project
Machine learning methods typically can only solve the tasks that they have been specifically trained to solve. They first adapt (train) a mathematical model on a number of examples and then apply the trained model. However, when trained models are faced with new situations, their performance drops significantly. In other words, these systems may not generalize well: they perform poorly in scenarios that are related to but different from those they were trained on. This poses a major obstacle to the effective and reliable use of machine learning in practical applications.

We need to find training methods and model structures that can learn to master new situations without forgetting previously learned knowledge to an excessive extent. Such models, which perform continual learning, are studied and developed in this project. We will focus on continual learning in situations where several devices cooperate and learn together, i.e., distributed learning. This is a situation of great practical interest but it can make generalization even more difficult to achieve. Using structured insights from mathematical analysis of these problems, we will develop and evaluate methods for continuous learning that can generalize.

Duties

• The PhD student will carry out research in the area of distributed machine learning
• The PhD student will actively contribute to setting up the research questions in their doctoral project
• They will take an active role in planning, implementing and, where necessary, modifying their research project
• The PhD student will gain advanced and up-to-date specialized knowledge in the area
• They will develop new theory and methods; and analyze the generalization performance of these methods
• The work as a doctoral student also includes writing scientific publications and presenting research results orally in various contexts such as project group meetings as well and international conferences

The main task of a doctoral student is to devote to the doctoral education, which includes both participation in research projects and doctoral education courses. The duties also include participating in teaching and other institutional tasks to a maximum of 20% of the working time.

Requirements
To meet the entry requirements for doctoral studies, you must

• hold a Master’s (second-cycle) degree inengineering physics, electrical engineering, machine learning, data science, computer science, applied mathematics or in a similar field, or
• have completed at least 240 credits in higher education in these fields, with at least 60 credits at Master’s level including an independent project worth at least 15 credits, or
• have acquired substantially equivalent knowledge in some other way. 

Additional qualifications
We are looking for candidates with:

• A strong interest in developing new theory and methods for machine learning
• Strong mathematical background
• Good proficiency in programming (preferably in Python)
• Good oral and written proficiency in English
• A structured, self-driven, independent approach to technical work and good collaboration skills
• Coursework or other experiences with the following subjects are valued: optimization, linear algebra, signal processing, probability, random processes, statistical machine learning and deep learning.  

Rules governing PhD students are set out in the Higher Education Ordinance chapter 5, §§ 1-7 and in Uppsala University's rules and guidelines .

About the employment
The employment is a temporary position according to the Higher Education Ordinance chapter 5 § 7. Scope of employment 100 %.Starting date 2025-09-01 or as agreed. Placement: Uppsala.

For further information about the position, please contact: Ayca Ozcelikkale, ayca.ozcelikkale@angstrom.uu.se

Please submit your application by 31st of March 2025, UFV-PA 2025/402.

Are you considering moving to Sweden to work at Uppsala University? Find out more about what it´s like to work and live in Sweden .

Website

https://uu.varbi.com/en/what:job/jobID:797262/type:job/where:39/apply:1

Research Field

Engineering

Education Level

Master Degree or equivalent

Research Field

Engineering

Years of Research Experience

None

Number of offers available

1

Company/Institute

Uppsala universitet

Country

Sweden

City

Uppsala

Geofield

City

Uppsala

Website

http://www.uu.se/en/about-uu/join-us/jobs/

• X (formerly Twitter)
• Facebook
• LinkedIn
• Whatsapp
• 
  More share options
  • E-mail
  • Pocket
  • Viadeo
  • Gmail
  • Weibo
  • Blogger
  • Qzone
  • YahooMail