Andrey Ogurtsov
Principal Biostatistician
Kyiv, Ukraine
Summary
👤
Senior🎓
Top SchoolAndrey Ogurtsov is a Principal Biostatistician based in Kyiv with over a decade of experience designing and analyzing phase I–IV and bioequivalence trials and leading statistical work for more than 20 clinical studies. He blends deep domain expertise in biostatistics with practical R programming and machine learning skills—contributing ML examples for the popular R interface to Keras, including a U-Net image segmentation implementation adapted for cross-platform use. Comfortable from CRF design and SAP authoring to regulatory interactions, Andrey combines classical inference and resampling methods with modern ML and deep learning (tf/keras, pytorch) to solve complex clinical-data problems. He also builds reproducible workflows, R packages, and Shiny apps, and leverages cloud, Docker, and out-of-core tools for large-scale analyses. Notably, his background in molecular biology research gives him an unusually strong bridge between wet-lab science and statistical modeling.
11 years of coding experience
Master's degree Biochemistry, Master's degree Biochemistry at Taras Shevchenko National University of Kyiv. Institute of BiologySpecialist (master) Finance and credit, Specialist (master) Finance and credit at Taras Shevchenko National University of Kyiv. Institute of Postgraduate Education
Английский
Stackoverflow
21reputation
4kreached
2answers
0questions
Github Skills (18)
convolutional-neural-networks10
r10
machine-learning10
keras10
image-segmentation10
segmentation10
neural-network10
computer-vision10
trainings9
modeling9
data-augmentation8
api6
dvi6
latex6
windows6
Programming languages (6)
Github contributions (5)
rstudio/keras3
Sep 2017 - Sep 2017
R Interface to Keras
Role in this project:
ML Engineer
ML EngineerContributions:5 commits, 3 PRs, 17 comments in 6 days
Contributions summary:Andrey primarily contributes to an example implementation of a U-Net architecture for image segmentation using the R interface to Keras. Their work involves defining the U-Net model, including convolutional layers, upsampling layers, and loss functions like dice_coef and bce_dice_loss. The user also addresses code formatting issues and adapts the example to run on both Windows and Linux operating systems, along with adding dataset information.
keras
statist-bhfz/data.table_doc_ru
Oct 2015 - Apr 2017
Contributions:18 commits, 17 pushes, 1 branch in 1 year 6 months
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.