Adaptive Tensor Network Decomposition for Multidimensional Machine Learning Theory and Applications

• Project/Area Number: 24K20849
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 61030:Intelligent informatics-related
• Research Institution: Institute of Physical and Chemical Research
• Principal Investigator: Qiu Yuning 国立研究開発法人理化学研究所, 革新知能統合研究センター, 基礎科学特別研究員 (30991145)
• Project Period (FY): 2024-04-01 – 2027-03-31
• Project Status: Granted (Fiscal Year 2024)
• Budget Amount: ¥4,680,000 (Direct Cost: ¥3,600,000、Indirect Cost: ¥1,080,000); Fiscal Year 2026: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000); Fiscal Year 2025: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000); Fiscal Year 2024: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000)
• Keywords: Tensor Network / Machine Learning / Tensor Decomposition / Deep Learning

Outline of Research at the Start

Learning from multi-relational and multi-modal data, often represented as high-order tensors, stands as one of the significant challenges within machine learning community. Tensor Network decomposition (TND) offers a promising solution to address the curse of dimensionality in these scenarios. However, the existing tensor network decomposition is limited by a specific topology structure, which makes it difficult to mine the potential data structure. This project intends to break through this limitation and develop adaptive TND-based machine learning methods, theory, and its applications.