Cortex inspired Deep Learning Algorithms and Applications on Knowledge Processing

2018 Fiscal Year Final Research Report

• Project/Area Number: 15H05327
• Research Category: Grant-in-Aid for Young Scientists (A)
• Allocation Type: Single-year Grants
• Research Field: Web informatics, Service informatics
• Research Institution: The University of Tokyo
• Principal Investigator: Nakayama Kotaro 東京大学, 大学院工学系研究科(工学部), 特任講師 (00512097)
• Project Period (FY): 2015-04-01 – 2019-03-31
• Keywords: Deep Learning / 人工知能 / 計算科学

Outline of Final Research Achievements

In this research project, we aimed to develop scalable deep learning algorithms using the latest brain science findings, and have conducted research on flexible knowledge processing mechanism. There are two points to achieve this mechanism; a general-purpose knowledge processing model applicable to various tasks (applications) and a calculation model optimized for parallel processing capable of processing large-scale data in real time is there. The basic & theoretical research part and applied research part have been progressed as scheduled, and the results have been published as international conference papers and domestic journals.

Free Research Field

AI、大規模計算

Academic Significance and Societal Importance of the Research Achievements

第三次AIブームを牽引しているDeep Learning技術は、計算時間や計算コストが大きな課題であった。本研究課題では，最新の脳科学の知見を活かしてより効率的かつ柔軟な知識処理機構を持つDeep Learning手法を実現することを目指して研究を進めてきた。予定どおり基礎研究と応用研究についての研究を進めることができ、国際会議・国内論文誌含め、多くの論文として対外発表することができた。