• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Statistical Learning with feature extraction and information integration of High-dimensional, large-scale, multi-domain data

Research Project

Project/Area Number 19H04071
Research Category

Grant-in-Aid for Scientific Research (B)

Allocation TypeSingle-year Grants
Section一般
Review Section Basic Section 60030:Statistical science-related
Research InstitutionTokyo Institute of Technology

Principal Investigator

Kanamori Takafumi  東京工業大学, 情報理工学院, 教授 (60334546)

Co-Investigator(Kenkyū-buntansha) 熊谷 亘  東京大学, 大学院工学系研究科(工学部), 特任助教 (20747167)
竹之内 高志  政策研究大学院大学, 政策研究科, 教授 (50403340)
松井 孝太  名古屋大学, 医学系研究科, 講師 (50737111)
川島 孝行  東京工業大学, 情報理工学院, 助教 (60846210)
武田 朗子  東京大学, 大学院情報理工学系研究科, 教授 (80361799)
Project Period (FY) 2019-04-01 – 2024-03-31
Project Status Completed (Fiscal Year 2023)
Budget Amount *help
¥17,290,000 (Direct Cost: ¥13,300,000、Indirect Cost: ¥3,990,000)
Fiscal Year 2023: ¥3,380,000 (Direct Cost: ¥2,600,000、Indirect Cost: ¥780,000)
Fiscal Year 2022: ¥2,470,000 (Direct Cost: ¥1,900,000、Indirect Cost: ¥570,000)
Fiscal Year 2021: ¥2,470,000 (Direct Cost: ¥1,900,000、Indirect Cost: ¥570,000)
Fiscal Year 2020: ¥2,470,000 (Direct Cost: ¥1,900,000、Indirect Cost: ¥570,000)
Fiscal Year 2019: ¥6,500,000 (Direct Cost: ¥5,000,000、Indirect Cost: ¥1,500,000)
KeywordsAI / データサイエンス / 機械学習 / 数理統計学 / 数理統計 / 機械学修 / 最適化 / 最適輸送 / 多ドメインデータ / 転移学習 / 情報転送
Outline of Research at the Start

本研究では,さまざまな形式で表現された高次元・大規模な多ドメインデータを用いて統計的学習を行うためのフレームワークを構築することを目的とします.今日のようなビッグデータ時代では,さまざまなデータドメインにおいて,サイズ,次元,表現形式の異なる複雑なデータを収集することが出来ます.これは一見すると data-rich な状態と言えます.しかし,それぞれのドメイン間の相互関係が不明なことが多く,その意味ではデータ量が増えるほどknowledge-poor な状態になってしまいかねません.このようなビッグデータ時代のパラドックスを打破するための研究を推進します.

Outline of Final Research Achievements

This study aims to construct a framework for statistical learning using high-dimensional and large-scale multi-domain data. In the era of big data, diverse and complex data with different sizes, dimensions, and representation formats can be collected across various domains. However, there exists a paradox wherein the relationships between domains are often unclear, leading to a knowledge deficit as data volume increases. To overcome this, it is crucial to extract and integrate features of data while considering the inter-domain relationships. This research focuses on formalizing this task with a focus on inter-domain relationships. It seeks to develop modeling techniques and machine learning algorithms for multi-domain data with heterogeneous structures, aiming to advance the theoretical understanding in this field.

Academic Significance and Societal Importance of the Research Achievements

本研究では,異なるデータサイズ,次元,タイプなどの多様なデータを活用し,予測,推論,構造推定など複数のタスクを行う学習アルゴリズムを,数学的な知見に基づいて提案,開発する.理論的解析により予測精度向上のためのパラメーター調整などが容易になり,飛躍的な性能向上が期待できる.理論的知見に基づくアルゴリズムの実装により,画像,音声,タグその他の情報を含むヘテロなデータからの関連性分析などの精度が大きく向上し,機械学習システムの安全性や信頼性を高める基盤を提供する。

Report

(6 results)
  • 2023 Annual Research Report   Final Research Report ( PDF )
  • 2022 Annual Research Report
  • 2021 Annual Research Report
  • 2020 Annual Research Report
  • 2019 Annual Research Report
  • Research Products

    (32 results)

All 2024 2023 2022 2021 2020 2019 Other

All Int'l Joint Research (3 results) Journal Article (11 results) (of which Int'l Joint Research: 4 results,  Peer Reviewed: 11 results,  Open Access: 4 results) Presentation (16 results) (of which Int'l Joint Research: 10 results,  Invited: 1 results) Book (1 results) Patent(Industrial Property Rights) (1 results)

  • [Int'l Joint Research] Max Planck Institute/Google Deepmind/Freie Universitat Berlin(ドイツ)

    • Related Report
      2023 Annual Research Report
  • [Int'l Joint Research] Technische Universitat Berlin/Berlin Institute for the Foundations(ドイツ)

    • Related Report
      2023 Annual Research Report
  • [Int'l Joint Research] University of Bristol(英国)

    • Related Report
      2022 Annual Research Report
  • [Journal Article] Denoising Cosine Similarity: A Theory-Driven Approach for Efficient Representation Learning2024

    • Author(s)
      T. Nakagawa, Y. Sanada, H. Waida, Y. Zhang, Y. Wada, K. Takanashi, T. Yamada, T. Kanamori
    • Journal Title

      Neural Networks

      Volume: 169 Pages: 226-241

    • DOI

      10.1016/j.neunet.2023.10.027

    • Related Report
      2023 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Learning Domain Invariant Representations by Joint Wasserstein Distance Minimization Learning Systems2023

    • Author(s)
      L. Andeol, Y. Kawakami, Y. Wadad, T. Kanamori, K. R. Muller, G. Montavon,
    • Journal Title

      Neural Networks

      Volume: 167 Pages: 233-243

    • DOI

      10.1016/j.neunet.2023.07.028

    • Related Report
      2023 Annual Research Report
    • Peer Reviewed / Int'l Joint Research
  • [Journal Article] Deep Clustering With a Constraint for Topological Invariance Based on Symmetric InfoNCE2023

    • Author(s)
      Y. Zhang, Y. Wada, H. Waida, K. Goto, Y. Hino,T. Kanamori
    • Journal Title

      Neural Computation

      Volume: 35 Issue: 7 Pages: 1288-1339

    • DOI

      10.1162/neco_a_01591

    • Related Report
      2023 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Estimating Density Models with Truncation Boundaries using Score Matching.2022

    • Author(s)
      S. Liu, T. Kanamori, and D. J. Williams,
    • Journal Title

      Journal of Machine Learning Research,

      Volume: 23 Pages: 1-38

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Open Access / Int'l Joint Research
  • [Journal Article] Uncertainty propagation for dropout-based Bayesian neural networks2021

    • Author(s)
      Mae Yuki、Kumagai Wataru、Kanamori Takafumi
    • Journal Title

      Neural Networks

      Volume: 144 Pages: 394-406

    • DOI

      10.1016/j.neunet.2021.09.005

    • Related Report
      2021 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Robust Label Prediction via Label Propagation and Geodesic <i>k</i>-Nearest Neighbor in Online Semi-Supervised Learning2019

    • Author(s)
      WADA Yuichiro、SU Siqiang、KUMAGAI Wataru、KANAMORI Takafumi
    • Journal Title

      IEICE Transactions on Information and Systems

      Volume: E102.D Issue: 8 Pages: 1537-1545

    • DOI

      10.1587/transinf.2018EDP7424

    • NAID

      130007686445

    • ISSN
      0916-8532, 1745-1361
    • Year and Date
      2019-08-01
    • Related Report
      2019 Annual Research Report
    • Peer Reviewed / Int'l Joint Research
  • [Journal Article] Spectral Embedded Deep Clustering2019

    • Author(s)
      Wada Yuichiro、Miyamoto Shugo、Nakagama Takumi、Andeol Leo、Kumagai Wataru、Kanamori Takafumi
    • Journal Title

      Entropy

      Volume: 21 Issue: 8 Pages: 795-795

    • DOI

      10.3390/e21080795

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed / Open Access / Int'l Joint Research
  • [Journal Article] Variable Selection for Nonparametric Learning with Power Series Kernels2019

    • Author(s)
      Matsui Kota、Kumagai Wataru、Kanamori Kenta、Nishikimi Mitsuaki、Kanamori Takafumi
    • Journal Title

      Neural Computation

      Volume: 31 Issue: 8 Pages: 1718-1750

    • DOI

      10.1162/neco_a_01212

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Numerical Study of Reciprocal Recommendation with Domain Matching2019

    • Author(s)
      K. Sudo, N. Osugi, T. Kanamori
    • Journal Title

      Japanese Journal of Statistics and Data Science

      Volume: 2 Issue: 1 Pages: 221-240

    • DOI

      10.1007/s42081-019-00033-3

    • NAID

      210000170705

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Model Description of Similarity-Based Recommendation Systems2019

    • Author(s)
      Kanamori Takafumi、Osugi Naoya
    • Journal Title

      Entropy

      Volume: 21 Issue: 7 Pages: 702-702

    • DOI

      10.3390/e21070702

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Risk bound of transfer learning using parametric feature mapping and its application to sparse coding2019

    • Author(s)
      Kumagai Wataru、Kanamori Takafumi
    • Journal Title

      Machine Learning

      Volume: 108 Issue: 11 Pages: 1975-2008

    • DOI

      10.1007/s10994-019-05805-2

    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Presentation] Robust VAEs via Generating Process of Noise Augmented Data2024

    • Author(s)
      H. Irobe, W. Aoki, K. Yamazaki, Y. Zhang, T. Nakagawa, H. Waida, Y. Wada, T. Kanamori,
    • Organizer
      IEEE International Symposium on Information Theory
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Open-World Learning Under Dataset Shift2024

    • Author(s)
      P. Srey, Y. Zhang, T. Kanamori,
    • Organizer
      IEEE Conference on Artificial Intelligence
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Local Acquisition Function for Active Level Set Estimation2023

    • Author(s)
      Y. Kokubun, K. Matsui, K. Kutsukake, W. Kumagai, T. Kanamori
    • Organizer
      NeurIPS 2023 Workshop on Adaptive Experimental Design and Active Learning in the Real World
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Towards Understanding the Mechanism of Contrastive Learning via Similarity Structure: A Theoretical Analysis2023

    • Author(s)
      H. waida, Y. Wada, L. Andeol, T. Nakagawa, Y. Zhang, T. Kanamori
    • Organizer
      European Conference on Machine Learning and Data Mining
    • Related Report
      2023 Annual Research Report
    • Int'l Joint Research
  • [Presentation] 平滑化全変動距離によるロバスト推定2023

    • Author(s)
      金森 敬文、横山 皓大、川島 孝行
    • Organizer
      統計関連学会連合大会
    • Related Report
      2023 Annual Research Report
  • [Presentation] 有界領域上における一致性のあるカーネル密度推定量の構成2023

    • Author(s)
      中川 匠、髙梨 耕作、金森 敬文
    • Organizer
      統計関連学会連合大会
    • Related Report
      2023 Annual Research Report
  • [Presentation] 局所探索型獲得関数に基づく能動的レベル集合推定法の提案2023

    • Author(s)
      國分裕太; 松井孝太; 沓掛健太郎; 熊谷亘; 金森敬文
    • Organizer
      情報論的学習理論ワークショップIBIS
    • Related Report
      2023 Annual Research Report
  • [Presentation] Fast Neural Architecture Search with Random Neural Tangent Kernel2023

    • Author(s)
      Keigo Wakayama; Takafumi Kanamori
    • Organizer
      情報論的学習理論ワークショップIBIS
    • Related Report
      2023 Annual Research Report
  • [Presentation] Deep Self-Supervised Learning of Speech Denoising from Noisy Speeches.2022

    • Author(s)
      Y. Sanada1, T. Nakagawa, Y. Wada, K. Takanashi, Y. Zhang, K. Tokuyama, T. Kanamori, T. Yamada,
    • Organizer
      INTERSPEECH 2022
    • Related Report
      2022 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Unified surrogate bounds for kernel-based contrastive unsupervised representation learning.2022

    • Author(s)
      和井田博貴; 和田裕一郎; Andeol Leo; 中川匠; Zhang Yuhui; 金森敬文
    • Organizer
      第25回情報理論的学習理論ワークショップ(IBIS2022)
    • Related Report
      2022 Annual Research Report
  • [Presentation] Denoising Cosine Similarity: A Theory-Driven Approach for Efficient Representation Learning2022

    • Author(s)
      中川匠; 眞田雄太郎; 和井田博貴; Zhang Yuhui; 和田裕一郎; 髙梨耕作; 山田知典; 金森敬文
    • Organizer
      第25回情報理論的学習理論ワークショップ(IBIS2022)
    • Related Report
      2022 Annual Research Report
  • [Presentation] Mode estimation on matrix manifolds: Convergence and robustness2022

    • Author(s)
      H. Sasaki, J. Hirayama, T. Kanamori,
    • Organizer
      The 25th International Conference on Artificial Intelligence and Statistics (AISTATS2022)
    • Related Report
      2021 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Robust modal regression with direct gradient approximation of modal regression risk.2020

    • Author(s)
      H. Sasaki, T Sakai, T. Kanamori,
    • Organizer
      The Conference on Uncertainty in Artificial Intelligence (UAI2020)
    • Related Report
      2020 Annual Research Report
    • Int'l Joint Research
  • [Presentation] A Unified Statistically Efficient Estimation Framework for Unnormalized Models2020

    • Author(s)
      M. Uehara, T. Kanamori, T. Takenouchi, T. Matsuda,
    • Organizer
      The 23rd International Conference on Artificial Intelligence and Statistics (AISTATS 2020)
    • Related Report
      2020 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Fisher Efficient Inference of Intractable Models.2019

    • Author(s)
      S. Liu, T. Kanamori, W. Jitkrittum, Y. Chen
    • Organizer
      The Neural Information Processing Systems (NeurIPS 2019), December 2019.
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research
  • [Presentation] Foundations of transfer learning and its application to multi-center prognostic prediction.2019

    • Author(s)
      K. Matsui, W. Kumagai, K. Kanamori, M. Nisikimi, S. Matsui, T. Kanamori
    • Organizer
      2019 WNAR/IMS/JR Annual Meeting
    • Related Report
      2019 Annual Research Report
    • Int'l Joint Research / Invited
  • [Book] データサイエンスと機械学習2022

    • Author(s)
      D.P.Kroese ほか著,金森 敬文 監訳
    • Total Pages
      416
    • Publisher
      東京化学同人
    • ISBN
      9784807920297
    • Related Report
      2022 Annual Research Report
  • [Patent(Industrial Property Rights)] 演算装置および学習済みモデル2021

    • Inventor(s)
      前 佑樹,金森 敬文
    • Industrial Property Rights Holder
      前 佑樹,金森 敬文
    • Industrial Property Rights Type
      特許
    • Filing Date
      2021
    • Acquisition Date
      2023
    • Related Report
      2023 Annual Research Report

URL: 

Published: 2019-04-18   Modified: 2025-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi