Kota Noda (24 years old, Japanese)

Bachelor of Engineering (the University of Tokyo 2018 ~ 2022)
Master of Engineering (the University of Tokyo 2022 ~ 2024)

Software Engineer at Techouse, Inc. 2022/09 ~

MD-GNN
abstructure

Competitions Expert (silver🥈 × 2, bronze🥉 × 1, Highest Rank: 1076)

[ Competitions ]
BirdCLEF 2022 50 of 807 teams 🥈(solo)
American Express - Default Prediction 242 of 4875 teams 🥈
Santa 2020 - The Candy Cane Contest 80 of 788 teams 🥉(solo)

[ Repositories ]
Kaggle_amex, Kaggle_birdclef2022

atnacup#13 (Retail AI hackathon 1st stage) 4 of 127 teams

nodematerial
Algorithm: highest 970
Heuristic: highest 414

I used to major in material simulation&informatics
mainly deal with

• Molecular dynamics (Masetr)
• Deep Learning: GNN mainly　(Masetr)
• Microscophic image analysis (Bachelor)

卒業論文 : 画像解析と深層学習に基づく水性塗料成膜・硬化機構解明
Bachelor's thesis : Elucidation of film formation and curing mechanism of water-based paint based on image analysis and deep learning

修士論文 : 深層学習による分⼦動⼒学の物理量時間発展モデル

原著論文

1. Prediction of potential energy profiles of molecular dynamic simulation by graph convolutional networks. Kota Noda, Yasushi Shibuta. Computational Materials Science. 229:112448, 2023.
2. Predicting long-term trends in physical properties from short-term molecular dynamics simulations using long short-term memory. Kota Noda, Yasushi Shibuta. J Phys Condens Matter. 36:385902, 2024.

深層⽣成モデルによる多結晶原⼦構造の特徴量抽出および復元 日本金属学会2022年秋期(第171回)講演大会　Kohei Sase, Kota Noda, Yasushi Shibuta

グラフニューラルネットワークを使用した深層学習モデルによる分子動力学シミュレーションの物理量予測 日本金属学会2023年秋期(第173回)講演大会　Kota Noda, Yasushi Shibuta

High-precision prediction of physical properties of molecular dynamic simulation using graph neural networks PRICM11 Kota Noda, Yasushi Shibuta

mainly: Python3, Ruby, C++, Ruby on Rails, Django
others: JavaScript, Go