Main
Brynjólfur Gauti Guðrúnar Jónsson
PhD Student of Statistics at the University of Iceland
I am a PhD student of statistics at the University of Iceland, researching spatially distributed extreme rainfall and methods for modeling data-level dependence using copulas. Prior to this, I have worked extensively with Bayesian hierarchical models, having applied them to forecasting parliamentary elections and age-dependent mortality as well as the modeling of COVID-19 disease spread in Iceland during the pandemic. I also manage metill.is, an Icelandic website where I use official data to inform the discourse on current matters.
Education
University of Iceland
PhD in Statistics
Reykjavík, Iceland
2027 (est)
Subject: Improved 21st century projections of sub-daily extreme precipitation by spatio-temporal recalibration
University of Iceland
M.S. in Statistics
Reykjavík, Iceland
2020
University of Iceland
B.S. in Psychology
Reykjavík, Iceland
2018
Thesis: Up high, down low, too slow: Effects of nature on directed attention differ between high and low spatial frequencies. Results were presented at the Biomedical and Health Sciences Conference at the University of Iceland.
Academic Experience
StanCon 2024
Oxford University
Oxford, England
2024
- I presented my research and talked about modeling with copulas in Stan Presentation slides
Royal Statistical Society 2024 International Conference
Doubletree Brighton Metropole
Brighton, England
2024
- I gave a talk on my PhD research into developing Gaussian copulas for large spatial fields. Presentation slides
PhD Student of Statistics
University of Iceland
Reykjavík, Iceland
2022 - present
- Researching improved 21st century projections of sub-daily extreme precipitation and how hierarchical Bayesian spatio-temporal models can be applied to bias-correct simulations from convection-permitting climate models using observed meteorological data.
Statistician
The Icelandic Heart Association
Reykjavík, Iceland
2018 - 2022
- High-dimensional statistical modeling of the effects of serum levels of proteins on heart failure and other outcomes.
Statistician
Directorate of Health
Reykjavík, Iceland
2020 - 2020
- Designed and implemented a Bayesian hierarchical generalized logistic growth model for short-term predictions of diagnosed cases at the start of the COVID19 pandemic.
Graduate Research Assistant
The Science Institute, University of Iceland
Reykjavík, Iceland
2019
- Performed preliminary research related to my MSc thesis and presented results at the 2019 Nordic Demographic Symposium in Reykjavík
Professional Experience
Data Scientist
Lucinity
Reykjavík, Iceland
2023 - 2024
- Exploring the use of privacy-enhancing technologies, machine learning and network analysis for detecting money laundering activities
- Graph neural networks, federated learning, simulating structured large-scale graph data
Statistician
The Icelandic Heart Association
Reykjavík, Iceland
2018 - 2022
- High-dimensional statistical modeling of the effects of serum levels of proteins on heart failure and other outcomes.
- Diverse data analysis projects to help with publications
- Writing software that helps coworkers with their reasearch
Statistical Consultant
The Statistical Consulting Center, School of Health Sciences, University of Iceland
Reykjavík, Iceland
2018 - 2021
- Assisted graduate students and academic researchers with data analysis and statistical testing
- Wrote methods chapters for research articles
- Did power calculations and helped plan methodology before data was gathered
Statistician
Directorate of Health
Reykjavík, Iceland
2020 - 2020
- Designed and implemented a Bayesian hierarchical generalized logistic growth model for short-term predictions of diagnosed cases at the start of the COVID19 pandemic.
- High speed data analysis used to gather actionable intelligence presented via interpretable graphics, tables and text
- Designed and developed a detailed dashboard
Researcher
Icelandic Center for Research - RANNÍS
Reykjavík, Iceland
2018 - 2018
- Designed and planned a research project and applied for funding
- Developed personalized risk calculator to predict patients’ risk of 30-day readmittance to psychiatric wards
- One of five projects nominated for the President’s Innovation Award
- Results were presented at the Biomedical and Health Sciences Conference at the University of Iceland.
Teaching Experience
Statistical Consulting
Teacher
University of Iceland, Reykjavík, Iceland
2020 - present
Applied Data Analysis
Teacher
University of Iceland, Reykjavík, Iceland
2020 - present
R Programming
Teacher
University of Iceland, Reykjavík, Iceland
2019 - present
Survival Analysis
Teaching assistant
University of Iceland, Reykjavík, Iceland
2019 - 2020