معرفی
Goki Yasuda is an Associate Professor at the Center for Data Science, Waseda University. His research spans statistical science, intelligent informatics, and machine learning, with specialized interests in label noise robustness, semi-supervised learning, and asymptotic analysis of algorithms. He teaches extensively in data science education, including courses like 'Statistics Literacy', 'Data Analysis with Python', and 'Modeling for Data Science' through Waseda's Global Education Center.
His research explores foundational machine learning challenges, including:
- Theoretical analysis of classification under label noise conditions
- Performance limits of semi-supervised learning with unlabeled data
- Bayesian inference for misspecified probabilistic models
- Algorithm optimization for real-world data imperfections
Publications demonstrate a strong emphasis on theoretical machine learning (75% of recent work) with secondary focus on applied signal processing for video coding (25% of earlier publications). Research consistently employs statistical methods to address data quality issues in learning systems.
He leads research projects including:
- 'Developing machine learning algorithms for low-quality data' (2022)
- 'Research on machine learning from data of various quality' (2019)
- 'Performance evaluation of prediction in half-ranked learning' (2018)
