Oliver Hall, Ph.D¶
ESA Research Fellow, Data Scientist, Astrophysicist¶
info¶
ojhall94@gmail.com
github.com/ojhall94
asteronomer.com
@asteronomer
(+31)(0)614227748
Leiden, NL
ORCID: 0000-0002-0468-4775
languages¶
Bilingual fluency in English and Dutch
programming¶
Fluent: Python, SQL
Competent: HTML, MATLAB
Python: NumPy, SciPy, pandas, scikit-learn, Stan, JAX, Pyro, PyMC, TensorFlow, Matplotlib, seaborn, pytest
Other: Jupyter, Google Colab, GitHub, Unix, Office365, ChatGPT
skills¶
- Bayesian Statistics
- AI/Machine Learning
- Predictive Modelling
- Open Source Development
- Data Visualisation
- Time-series analysis
- Communication
- Leadership, Mentoring
- Critical Thinking
- Problem Solving
- Project Management
awards + hons¶
ESA Research Fellowship:
Highly competitive fellowship for independent research
Associate Fellow (AFHEA):
Formal acknowledgement of teaching expertise
Royal Society Partnership Grant:
Competitive £3000 grant for school outreach programme
Main body:¶
PhD in Astrophysics with a focus on Bayesian statistics, machine learning and high performance computational methods. Track record of finding creative data-driven solutions to open problems and delivering actionable scientific results. Thrives in a fast-paced, deadline-driven collaborative environment. 7 years experience in research of which 3 post-doctoral. Seeking new challenges and growth in data and industry.
Experience:¶
2020 to now - ESA Research Fellow - European Space Research and Technology Centre, NL
- Applying Fourier Analysis, Gaussian Processes, Hierarchical Bayesian modelling to time-series data using PyMC3, Pyro, JAX.
- Providing statistical and scientific expertise to ESA staff scientists.
- Designing, executing and disseminating novel research and its outcomes.
2020 - Freelance Software Developer - NumFOCUS, US
Projects:¶
Automated all-sky survey of stellar rotation in galactic structures
- Built open source program automating measurement of stellar rotation in 213$\,$GB of data.
- Built a hierarchical latent variable model reporting 790 stellar membership probabilities of galactic structures used for follow-up observations and analysis by scientific community.
Advanced hierarchical models of stellar rotation in time-series data
- Used hierarchical models, Gaussian processes to measure rotation in 94 stellar periodograms.
- Performed model rejection between two hypotheses in 5-dimensional parameter space, providing critical new evidence that a contested theory of stellar evolution was 98\% more likely than its alternative.
Modelling covariances of stellar properties in large standard-candle populations
- Developed generative hierarchical latent variable models of population of >5000 stars.
- Delivered detailed information about covariances in the population properties, leading to a 25\% precision improvement of distance measurements calibrated using this population.
education¶
2016 to 2020 - Ph.D. in Physics \& Astronomy - University of Birmingham, UK
- Developed novel Bayesian Models to determine stellar population properties.
- Lead development on two modules of open source Python package Lightkurve, with >250k downloads and >400 citations.
2012 to 2016 - M.Sci. in Physics \& Astrophysics - 1$^{st}$ w. Hons. - University of Birmingham, UK
2006 to 2012 - Gymnasium - 10 9s incl. Maths, Physics, English - Gemeentelijk Gymnasium Hilversum, NL
other experience¶
- Collaborating: worked on projects and in consortia in teams ranging from 3 to 80+ researchers.
- Speaking: expert speaker for >20 academic conferences, seminars, public outreach events.
- Writing: (co-)authored 18 scientific and 14 popular-science articles, receiving 850+ citations.
- Data presentation: designed readable graphs, posters, presentations for experts and public.
- Peer review: expert panel member for 2 scientific journals and 2 NASA funding programs.
- Teaching/Mentoring: (co-)supervisor for 4 master's students, run knowledge workshops.
- Organising: run 3 conferences, summer student program, fortnightly working group meetings.