Scholar
Nontawat Charoenphakdee
Google Scholar ID: sEFoFbgAAAAJ
Researcher, Preferred Networks
Citations & Impact
Citations
1,313
H-index
12
i10-index
14
Publications
20
Co-authors
23
list available
Publications
Resume (English only)
Academic Achievements
- 2025: Paper 'Calm Composite Losses: Being Improper Yet Proper Composite' accepted to AISTATS 2025; 2023: Paper 'Is the Performance of My Deep Network Too Good to Be True? A Direct Approach to Estimating the Bayes Error in Binary Classification' accepted to ICLR 2023; 2022: Multiple papers accepted to NeurIPS 2022 workshops; 2022: Paper 'Towards universal neural network potential for material discovery applicable to arbitrary combination of 45 elements' accepted to Nature Communications; Recipient of Google Ph.D. Fellowship in Machine Learning (2020) and Dean’s Award for outstanding research achievement from the University of Tokyo.
Research Experience
- Currently a researcher at Preferred Networks, focusing on machine learning for healthcare and quantum chemistry domains.
Education
- Completed a PhD under the supervision of Prof. Masashi Sugiyama in the Sugiyama-Yokoya-Ishida laboratory, Department of Computer Science, The University of Tokyo. His doctoral thesis was titled 'Theory and algorithms of machine learning with rejection based on loss function perspective.'
Background
- Research interests include machine learning algorithm analysis and design, specifically in loss functions, evaluation metrics, learning with a reject option, weakly supervised learning, domain adaptation, anomaly detection, and uncertainty quantification. Applications include healthcare data analysis, neural network potentials for quantum chemistry, speech signal processing, computer vision, and natural language processing.
Miscellany
- Personal website powered by Jekyll with al-folio theme, hosted by GitHub Pages.
