Daniel Homola
Founder at Stealth
Stony Stratford, England, United Kingdom
Summary
🤩
Rockstar🎓
Top SchoolDaniel Homola is a machine learning engineer, researcher and founder with a decade of experience translating cutting-edge ML research into production products across healthcare, drug discovery and geospatial risk. He holds a PhD in Biomedical Machine Learning from Imperial College London and has led ML efforts at companies like Exscientia, Tractable and Sensyne Health, including regulatory-grade Bayesian models deployed in hospitals and a published multimodal disaster-prediction model. Daniel is also an active open-source contributor—authoring and improving BorutaPy, a widely used Python implementation of the Boruta feature-selection method—and has repeatedly built core ML frameworks and secure data platforms for sensitive biomedical datasets. Equally comfortable with research papers and shipping production systems, he combines deep domain expertise in biological and clinical data with entrepreneurial experience founding startups to unlock value from messy, open-ended data.
10 years of coding experience
11 years of employment as a software developer
Doctor of Philosophy (PhD), Biomedical Machine Learning, Doctor of Philosophy (PhD), Biomedical Machine Learning at Imperial College London
Github Skills (11)
scikit-learn10
algorithm10
algorithms10
machine-learning10
python10
implement10
feature-selection10
scikit10
numpy9
random-forest9
data-science8
Programming languages (5)
Github contributions (5)
scikit-learn-contrib/boruta_py
Jan 2016 - Jan 2021
Python implementations of the Boruta all-relevant feature selection method.
Role in this project:
ML Engineer
ML EngineerContributions:4 releases, 2 reviews, 52 commits in 5 years
Contributions summary:Daniel implemented and refined the BorutaPy feature selection method, a Python implementation of the Boruta algorithm. Their contributions included modifying the core algorithm for improved performance and adding features like a two-step correction and percentile-based thresholding. The user also refactored and unified the code, addressing dependencies and incorporating features from previous versions. Additionally, they added example files and updated documentation to improve usability and maintainability of the library.
implementationspythonmethodscikit-learnselection
danielhomola/CorrMapper
Oct 2017 - Nov 2017
Contributions:6 commits, 3 pushes, 1 branch in 1 month
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.