Early Stage Researcher, Algorithmic Data Analysis

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 101034307.

Tensor and matrix decompositions for analysis of multimodal data in neuroscience

Doctoral Programme in Science, Technology and Computing (SCITECO)


Prof. Pauli Miettinen, primary supervisor
Prof. Jussi Tohka
Prof. Ville Leinonen
Dr. Outi-Maaria Palo-oja


School of Computing
Algorithmic Data Analysis group

This research project needs the participation of people with passion for excellence in research, innovation, and societal impact. We are looking for candidates who seek to contribute to brain health innovation research with a keen eye on the application of research knowledge to the problems and challenges of society and industry. Ideal candidates want to open their minds to intersecting and complementary fields of science, seeking a multi/interdisciplinary approach in research.

The academic profile of the candidate should be in data science or computer science, with the idea of complementing this background throughout PhD studies. The candidate must have a good understanding of linear algebra and knowledge of basic matrix decomposition methods (e.g. SVD or NMF). Previous knowledge of neuroscience is not necessary.

Overall, the candidate must be willing to combine complementary areas of knowledge in neurosciences, data sciences, social sciences, applied physics and/or law to complete the thesis project. Moreover, the candidate is required to have good methodological skills, especially in longitudinal analyses. As far as soft skills are concerned, we welcome candidates with a proactive, collaborative attitude, and who are enthusiastic about building, creating, and working in teams. Excellent candidates desire to participate in activities concerning multi- and interdisciplinary ageing research, and are willing to spend some time abroad with our international partners and outside the university in our partner institutions.

Data science, especially linear and multilinear algebra methods, neuroscience and innovation management.

Matrix and tensor decompositions are well-established methods for finding patterns in data. Recently there has also been increasing use of related techniques in multi-modal data that cannot be expressed as a single tensor or matrix (e.g. joint data from genetics, imaging, laboratory tests, behaviour and lifestyle). In this project, the goal is to develop new methods for multimodal matrix and tensor factorizations that can be applied for multiple data-analysis tasks in neuroscience.

The method development will be targeted to the studies of idiopathic normal pressure hydrocephalus (iNPH) and Alzheimer’s disease (AD), where a long-term goal is to develop a diagnostic and treatment protocol of iNPH optimized for prediction of clinical outcome using unique brain biopsy, imaging and genetic derived data sets. The work is based on existing methods, such as PARAFAC2 tensor decompositions and shared subspace matrix factorizations, but can involve development of novel methods that take into account the specifics of the application area. The successful candidate should have knowledge of matrix and tensor decompositions and linear algebra as well as algorithm design and computer science. The candidate should be interested in learning about clinical neuroscience related to iNPH and AD.

The candidate can also suggest their own topic or modify this topic in their motivation letter.


Publications related to the PhD topic:

  • Sanjar Karaev & Pauli Miettinen: Algorithms for Approximate Subtropical Matrix Factorization. Data Mining and Knowledge Discovery 33(2), 2019, pp. 526–576.
  • Saskia Metzler & Pauli Miettinen: Clustering Boolean Tensors. Data Mining and Knowledge Discovery 29(5), 2015, pp. 1343–1373.
  • Pauli Miettinen: On Finding Joint Subspace Boolean Matrix Factorizations. Proc. 2012 SIAM International Conference on Data Mining (SDM2012), 2012, pp. 954–965.
  • Moradi E, Pepe A, Gaser C, Huttunen H, Tohka J. Machine learning framework for early MRI-based Alzheimer’s conversion prediction in MCI subjects. Neuroimage. 2015 Jan 1;104:398-412.
  • Leinonen V, Kuulasmaa T & Hiltunen M. iNPH – the mystery resolving. EMBO Mol Med. 2021 Mar 5; 13(3):e13720

Research projects related to the PhD topic:

  • North-Savo AI-Hub is an ongoing project that aims at facilitating the use of novel technologies in social and health care data. The particular focus of the Algorithmic Data Analysis research group in the project is to develop novel methods for clinical data analysis and for privacy preserving data analysis of health care data.
  • The “Predictive brain image analysis” project funded by the Academy of Finland (PI Tohka) aims to develop new machine learning algorithms for predictive modeling of brain diseases, with an emphasis on the detection of very early signs of neurodegenerative diseases.

This PhD research topic covers the following complementary areas of knowledge: data science, especially linear and multilinear algebra methods; machine learning and image analysis; clinical neuroscience, neuro-innovation.

The PhD project will be co-supervised by a team of three scientists with expertise in data science, machine learning, biomedical image analysis, clinical neuroscience and innovation management.

To advance multi/interdisciplinary collaboration, three Summer Schools will be organised jointly by the Neuro-Innovation supervisors and non-academic partners. In these, multi-/interdisciplinarity and intersectoral exchange will be implemented via hands-on interaction between PhD students, supervisors, partners and other stakeholders.

  • Neuro-ethics and patient rights, 3 ECTS credits. Research ethics, patient rights, data security.
  • Neuro-data Hackathon, 3 ECST credits. Open and big data.
  • Neuro-Innovation Living Lab, 3 ECST credits. Entrepreneurial processes and commercialisation paths.

Virtual Platform for multi/interdisciplinary interaction will connect all PhD students in this programme.

The academic partner organizations (see the Neuro-Innovation webpage) will collaborate closely with us in PhD training and students are strongly encouraged to include a secondment and visits with these partners in their studies. During shorter visits (1-4 weeks), you will learn more about research and methods and build international networks. During secondments, you will work on their research project under the supervision of the co-supervisor from the hosting organisation and utilise their infrastructure. You can also attend courses, seminars, and other events in the host organisations.

The Societal Impact Board of this PhD programme with 14 intersectoral partners will collaborate closely with us in PhD training, for instance on the following activities:

  • Neuro-Innovation Talent Hub: monthly gathering with special guests (e.g., researchers, professionals, business experts, stakeholders) and discussions about research topics and career prospects.
  • Neuro-Innovation Boot Camp: yearly competition concerning the utilisation of research results.

Apply the position

High quality research and publications of excellence are a way to solve scientific and societal problems and challenges. This vision is shared by the collaborating partners in this doctoral program, and involvement with various stakeholders will allow the pursuit of a goal with potential for brain health innovation. The University of Eastern Finland pursues societal impact that goes beyond academia focusing on the transformation of society, leading to fairer and more diverse societies, where inclusive social development and welfare are enhanced.

The objective of the proposed research is to advance the analysis techniques of multimodal data, especially those pertaining to iNPH and Alzheimer’s disease. Better analysis can allow earlier detection and better treatment of the conditions.

The UEF policy on gender equality and equal opportunities is based on Finnish legislation and the values of the university. The goal of gender equality and equal opportunities at the university is to identify and prevent expressions, structures and functions that maintain or increase inequality and to promote gender equality and equal opportunities at all levels. The university has a Gender Equality and Equal Opportunities Programme, which describes the measures intended to implement and promote gender equality and equal opportunities among staff and students. The university takes an active approach to promoting equal opportunities and acting against discrimination.

The University of Eastern Finland is the most multidisciplinary university in Finland. We are home to 15,500 students and 2,700 staff members. Our research is ranked among the best in the world in several fields (inc. forest sciences). We generate research-based knowledge and make it openly accessible for the benefit of all. UEF stands for action with impact that is relevant today and tomorrow. To learn more about our university please visit our website at www.uef.fi/en.