Nontawat Charoenphakdee

นนทวัฒน์ เจริญภักดี (นัท), チャロエンパックディーノンタワット（ナット）

Selfie taken in Tokyo, November 2018.

I am a PhD student under the supervision of Prof. Masashi Sugiyama in Sugiyama-Sato-Honda laboratory, Department of computer science, The University of Tokyo.

Research interests: I am interested in machine learning algorithm analysis and design in general. More specifically, my current research topics include:

Loss function
Learning with reject option
Weakly-supervised learning
Domain adaptation

Links: Publications, CV

News

Nov 12, 2019	I gave an oral presentation: “Symmetric Loss and Its Applications to Learning under Label Noise” in Focus Program on Data Science and Optimization: Conference on Data Science at The Fields Institute, Toronto, Canada. The content was about our ICML2019 and EMNLP-IJCNLP2019 papers.
Nov 10, 2019	Our paper: “Classification from Triplet Comparison Data” has been accepted to Neural Computation.
Oct 21, 2019	I wrote about the project I did in PFN as an intern which is “Open-set Few-shot Speaker Identification”, in Preferred Networks Research Blog. I am really grateful to my mentors: Yuta Tsuboi and Katsuhiko Ishiguro, as well as other members in the team. Also, it is my pleasure to meet other interns and to have a chance to enjoy the internship program together!
Sep 25, 2019	I gave a poster presentation of our paper “On Symmetric Losses for Learning from Corrupted Labels” in IIT-RIKEN Joint Workshop, Genova, Italy.
Sep 4, 2019	Our paper: “On the Calibration of Multiclass Classification with Rejection” has been accepted to NeurIPS2019.
Aug 12, 2019	Our paper: “Learning Only from Relevant Keywords and Unlabeled Documents” has been accepted to EMNLP-IJCNLP2019.
Jul 12, 2019	I gave a poster presentation of our paper “On Symmetric Losses for Learning from Corrupted Labels” in SNL2019 Workshop, Tokyo, Japan.
Jun 2, 2019	I gave a poster presentation of our paper “On Symmetric Losses for Learning from Corrupted Labels” in PAIR-AIP Workshop, Taipei, Taiwan.
Apr 22, 2019	Our two papers: “On Symmetric Losses for Learning from Corrupted Labels” and “Imitation Learning from Imperfect Demonstration” have been accepted to ICML2019.
Dec 22, 2018	Our paper: “Positive-Unlabeled Classification under Class Prior Shift and Asymmetric Error” has been accepted to SDM2019.
Nov 1, 2018	Our paper: “Unsupervised Domain Adaptation Based on Source-guided Discrepancy” has been accepted to AAAI2019.