SATO Naoki, a 2nd-year master’s student at the Mathematical Optimization Laboratory (Professor IIDUKA Hideaki), won the the 5th Young Researcher’s Essay Award for Excellence of Tome R&D Inc. (prize money: 100,000 yen).
The award ceremony was held during the 2nd Artificial Intelligence Advanced Technology Lecture held at Kyoto Research Park, Buzz Hall on November 9.
The essay focuses on the stochastic noise implicit in optimization algorithms used in machine learning and examines the relationship between this noise and the generalization performance of deep learning models both theoretically and experimentally.
Title:
Relationship between search direction noise of continuous optimization algorithm and generalization performance of deep learning models