Yoshihiko Ozaki
Visiting Academic Staff at The University of Osaka
Tokyo, Japan
Summary
🤩
Rockstar🎓
Top SchoolYoshihiko Ozaki is a software engineer and researcher with eight years' experience blending AutoML, black-box optimization, and production engineering, currently serving as Visiting Academic Staff at the University of Osaka and an engineer on Preferred Networks' AutoML team. He holds a Ph.D. in Policy and Planning Sciences and is an active Optuna committer known for fixing and extending the TPE sampler and implementing MOTPE, contributing to multi-objective hyperparameter optimization in a widely used open-source framework. His background spans applied research at AIST—publishing award-winning surveys and methods used in materials science—and product leadership building OptunaHub as a feature-sharing platform. Comfortable moving between research and production, he also has industry experience accelerating large-scale builds at Google and developing optimization-driven systems for games and ML products. A detail-oriented engineer, he frequently improves core algorithm robustness, type hints, and compatibility to make advanced optimization techniques more reliable in practice.
9 years of coding experience
8 years of employment as a software developer
Doctor of Philosophy - PhD, Policy and Planning Sciences, Doctor of Philosophy - PhD, Policy and Planning Sciences at University of Tsukuba
Japanese, English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (5)
hyperparameter-optimization10
machine-learning10
back-end-development10
python10
numpy10
Programming languages (8)
Github contributions (5)
optuna/optuna
Apr 2020 - Jul 2022
A hyperparameter optimization framework
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:3 releases, 265 reviews, 65 commits in 2 years 3 months
Contributions summary:Yoshihiko primarily focused on modifying and fixing code related to the TPE sampler within the Optuna framework. They addressed issues with conditional parameters and made several corrections to the `split_observation_pairs` and `sample_independent` functions. The user also improved type hints and made minor fixes to enhance code compatibility, indicating a strong focus on the core functionality and robustness of the hyperparameter optimization process. Moreover, they were actively involved in the implementation of the MOTPE algorithm, which suggests contributions toward the development of multi-objective optimization capabilities.
hyperparameter-optimizationpythonmachine-learningparalleldistributed
y0z/optunahub
Jun 2024 - Jun 2026
Python library to use packages in OptunaHub Registry
Contributions:97 pushes, 47 branches in 2 years
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.