Sort by
Refine Your Search
-
Listed
-
Category
-
Employer
-
Field
-
Transformer-Based Foundation Model for DNA Methylation in Longitudinal Cohorts.” The focus is on developing next-generation AI models for the analysis of DNA methylation. Using longitudinal data from, among
-
longitudinal cohorts, SATSA and Betula, integrating established dementia biomarkers with inflammatory, metabolic, and genetic data using advanced statistical modeling and data-driven methods. The doctoral
-
principled new models and methods, for modern machine learning problems. Machine learning recently has been largely advanced by differential equation-based frameworks, such as generative diffusion models
-
application! We are looking for a PhD student for sustainable and resource-efficient machine learning. Your work assignments Machine learning has recently advanced through scaling model sizes, training budgets
-
well as physically-based hydrological model development. The principal supervisor will be Ylva Sjöberg at Umeå University, and the research involves an interdisciplinary team of collaborators at Gothenburg University
-
. 10 PhD vacancies are available in six universities. For more information, regularly check the individual vacancy pages of the universities listed on the project website [https://heritour.eu/ ]. About
-
scaling model sizes, training budgets, and datasets; often at substantial computational and environmental costs. This PhD project targets sustainable and resource-efficient machine learning with a focus on
-
model for large scale assessments and projections of the land-water carbon cycle to variation in climate conditions. The detailed direction of the PhD studies will be discussed and decided jointly with
-
and CH4) from headwaters, and use of machine learning and process-based model for large scale assessments and projections of the land-water carbon cycle to variation in climate conditions. The detailed
-
transformations. The project investigates a hybrid approach that combines deep learning with grammatical inference to develop models that are interpretable, efficient, and mathematically verifiable while leveraging