Romain Tavenard
Professeur Des Universités at tslearn
Brittany, France
Summary
🤩
Rockstar🎓
Top SchoolRomain Tavenard is a professor of computer science at Université Rennes 2 and a specialist in machine learning for environmental time series, with a PhD from Université de Rennes 1 and engineering training from École Centrale de Lyon and ENS Rennes. He created and continues to lead development of tslearn, a widely used open-source toolkit for time-series ML, and has contributed optimized algorithms to prominent libraries such as PythonOT/POT (notably refactoring the 1D Earth Mover’s Distance for sparse representations). Over a decade in research and academia, he blends rigorous theoretical work on kernel and alignment methods with practical engineering to improve performance and usability. Based in Brittany, he balances a strong publication and teaching record with hands-on open-source development, often surfacing mathematical insights directly in code and documentation.
10 years of coding experience
9 years of employment as a software developer
Diplôme d'ingénieur, Computer Science, Diplôme d'ingénieur, Computer Science at École Centrale de LyonDoctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Université de Rennes IComputer Science, Computer Science at École normale supérieure de Rennes
Github Skills (13)
kernel-methods10
machine-learning10
time-series10
python10
clustering10
numpy10
data-analysis10
data-science9
cython9
sparse9
scikit8
scikit-learn8
numba7
Programming languages (7)
Github contributions (5)
tslearn-team/tslearn
May 2017 - Jan 2023
The machine learning toolkit for time series analysis in Python
Role in this project:
Data Scientist & ML Engineer
Data Scientist & ML EngineerContributions:22 releases, 58 reviews, 1175 commits in 5 years 9 months
Contributions summary:Romain contributed to the development of kernel-based machine learning methods for time series analysis within the tslearn toolkit, as evidenced by code changes to the clustering and metrics modules. The commits reveal improvements in the implementation of the Global Alignment Kernel k-means (GAK) algorithm, showcasing expertise in kernel methods, time series analysis, and clustering techniques. The user focused on enhancing the accuracy and performance of the algorithm by adding various parameters (such as n_jobs for parallelization) and ensuring robust code.
dtwdynamic-time-warpingforecastingpythontime-series-analysis
PythonOT/POT
Jun 2019 - May 2020
POT : Python Optimal Transport
Role in this project:ML Engineer
ML EngineerContributions:28 commits, 4 PRs, 22 comments in 10 months
Contributions summary:Romain primarily contributed to the `POT` repository, focusing on the implementation and optimization of the Earth Mover's Distance (EMD) algorithm for 1D measures. Their work included refactoring the EMD1d implementation to utilize a sparse matrix representation, leading to performance improvements. The user also added documentation, implemented Minkowski metrics, and incorporated relevant mathematical formulas within the functions.
pythonot-mapping-estimationnumbanumerical-optimizationpattern-recognition
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.