Pavlin Poličar
Teaching Assistant at American Slovenian Education Foundation: ASEF
Slovenia
Summary
🤩
Rockstar🎓
Top SchoolPavlin Poličar is a PhD candidate and teaching assistant at the University of Ljubljana with 12 years of experience at the intersection of bioinformatics and machine learning, specializing in unsupervised representation learning and temporal embeddings for large-scale biomedical data. He combines academic research with hands-on software engineering—authoring and optimizing open-source tools such as openTSNE (notable for extensible, parallel t-SNE implementations) and VERA, and contributing to the Orange data-mining toolkit. His work spans method development, Cython performance tuning, and practical tooling that makes dimensionality reduction scalable and reproducible. Pavlin has collaborated internationally, including research fellowships at Harvard and visits to institutions in Padova and Florida, reflecting a strong network in computational biology. Outside academia he pursues photography, travel, and hiking, interests that inform his visual and exploratory approach to data visualization.
12 years of coding experience
4 years of employment as a software developer
FRI, PhD, Bioinformatics & Machine Learning, FRI, PhD, Bioinformatics & Machine Learning at University of LjubljanaDiocesan Classical Gymnasium, Diocesan Classical Gymnasium at St. Stanislav's Institution
English, French, Slovenian
Stackoverflow
5,438reputation
257kreached
72answers
53questions
if-statement
top-5%
webpack
top-5%
javascript
top-5%
babeljs
top-5%
Github Skills (23)
python10
machine-learning10
tsne10
cython10
javascript9
algorithm9
code-optimization9
algorithms9
webpack9
if-statement9
optimisation9
numerical-optimization9
optimization9
babeljs9
scikit-learn7
Programming languages (9)
Github contributions (5)
pavlin-policar/openTSNE
Jun 2018 - Dec 2022
Extensible, parallel implementations of t-SNE
Role in this project:
ML Engineer
ML EngineerContributions:7 releases, 23 reviews, 587 commits in 4 years 6 months
Contributions summary:Pavlin significantly contributed to the t-SNE implementation by adding new functionalities, including the ability to utilize different gradient methods within the optimization process, such as a custom implementation, Barnes-Hut, or interpolation based methods. They also added the option to use precomputed nearest neighbors and made several performance-related improvements to the Cython code for better efficiency. The user also refactored and renamed the cython functions for more consistency and added a gradient descent optimization with callback.
implementationsdimensionality-reductionparallelmulticoremachine-learning
conda-forge/opentsne-feedstock
Apr 2019 - Nov 2024
A conda-smithy repository for opentsne.
Contributions:1 review, 54 PRs, 39 pushes in 5 years 8 months
opentsnepythoncondasmithy
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.