Journal Paper

• Naoki Hayashi, SUMIO WATANABE. Upper Bound of Bayesian Generalization Error in Non-Negative Matrix Factorization, NEUROCOMPUTING, Elsevier Science Publishers, Vol. 266, pp. 21-28, Nov. 2017.
• Koda, N., SUMIO WATANABE. Interpretation method of nonlinear multilayer principal component analysis by using sparsity and hierarchical clustering, Proceedings - 2016 15th IEEE International Conference on Machine Learning and Applications, ICMLA 2016, pp. 1063-1066, 2017.

Book

• SUMIO WATANABE. 学習の代数的理論（人工知能学大辞典）, 共立出版, July 2017.

International Conference (Reviewed)

• Naoki Hayashi, SUMIO WATANABE. Tighter Upper Bound of Real Log Canonical Threshold of Non-negative Matrix Factorization and its Application to Bayesian Inference, 2017 IEEE Symposium Series on Computational Intelligence, Nov. 2017.
• Masahiro Koujima, SUMIO WATANABE. Phase Transition Structure of Variational Bayesian Nonnegative Matrix Factorization., 26th International Conference on Artificial Neural Networks 2017, Sept. 2017.

International Conference (Not reviewed / Unknown)

• SUMIO WATANABE. Difference between Bayes Cross Validation and WAIC for Conditional Independent Samples, The Tenth Workshop on Information Theoretic Methods in Science and Engineering, Sept. 2017.

Domestic Conference (Not reviewed / Unknown)

• Kouya Ohashi, SUMIO WATANABE. 変化点検出問題におけるベイズ検定統計量の導出と検出力の実験的考察, 第116回数理モデル化と問題解決(MPS)研究会, Dec. 2017.
• SUMIO WATANABE. 学習理論よ何処へ, 情報論的学習理論と機械学習研究会, Nov. 2017.
• Naoki Hayashi, SUMIO WATANABE. 確率行列分解の実対数閾値とBayes学習への応用, 情報論的学習理論と機械学習研究会, Nov. 2017.
• SUMIO WATANABE. 情報量規準とクロスバリデーションの同じ点と異なる点, 2017年統計サマーセミナー, Aug. 2017.

Research Highlight

• Masahiro Koujima, SUMIO WATANABE. ICANN Best Paper Award. 2017-2017.
• SUMIO WATANABE. Highly Cited Paper: Asymptotic equivalence of Bayes Cross Validation and WAIC. 2017-. 被引用回数上位１％論文に選ばれました。Sumio Watanabe, Asymptotic Equivalence of Bayes Cross Validation and Widely Appli....
  

[ Save as BibTeX ]    [ Paper, Presentations, Books, Others, Degrees: Save as CSV ]    [ Patents: Save as CSV ]