機械学習による超高次元データの統計的マッチングとマイクロアレイ解析への応用

2013 Fiscal Year Annual Research Report

• Project/Area Number: 23700165
• Research Institution: Nagoya Institute of Technology
• Principal Investigator: 竹内 一郎 名古屋工業大学, 工学(系)研究科(研究院), 准教授 (40335146)
• Keywords: 機械学習 / マイクロアレイ / 多変量検定

Research Abstract

平成２５年度は平成２４年度までに構築したマイクロアレイデータのための多変量検定統計の効率的な実装を行い、その性能を数値実験により検証した。さらに, 本年度は主双対ギャップの概念を利用した帰無分布の推定の高速化とGreedyアルゴリズムを用いた高速計算アルゴリズムを構築した. 本研究の関連研究の成果4件をInternational Conference on Machine Learning (ICML)とNeural Information Processing Systems (NIPS)にて発表した.

Research Products

• [Presentation] Global solver and its efficient approximation for variational bayesian low-rank subspace clustering (2013)
  • Author(s): S. Nakajima, A. Takeda, S. D. Babacan, M. Sugiyama and I. Takeuchi
  • Organizer: The 27th Annual Conference on Neural Information Processing Systems
  • Place of Presentation: Reno, USA
  • Year and Date: 20131205-20131211
• [Presentation] Parametric Task Learning (2013)
  • Author(s): I. Takeuchi, T. Hongo, M. Sugiyama and S. Nakajima
  • Organizer: The 27th Annual Conference on Neural Information Processing Systems
  • Place of Presentation: Reno, USA
  • Year and Date: 20131205-20131211
• [Presentation] Infinitesimal annealing for training semi-supervised support vector machines (2013)
  • Author(s): K. Ogawa, Imamura, I. Takeuchi, M. Sugiyama
  • Organizer: The 30th International Conference on Machine Learning
  • Place of Presentation: Atlanta, USA
  • Year and Date: 20130616-20130621
• [Presentation] Safe screening of non-support vectors in pathwise SVM computation (2013)
  • Author(s): K. Ogawa, Y. Suzuki, I. Takeuchi
  • Organizer: The 30th International Conference on Machine Learning
  • Place of Presentation: Atlanta, USA
  • Year and Date: 20130616-20130621