Mon, Mar 6 PM Theory of machine learning
Chair: Taiji Suzuki
13:30 - 16:15 |
| (1) |
13:30-14:00 |
Tighter generalization bounds for matrix completion based on non-negative matrix factorization with norm constraints |
Ken-ichiro Moridomi, Kohei Hatano, Eiji Takimoto (Kyushu Univ.) |
| (2) |
14:00-14:30 |
Exact calculation of grid-structured Markov random field model by corner transfer matrix method |
Tomoharu Yoshida, Kazuho Watanabe, Kyoji Umemura (Toyohashi Univ. of Tech.) |
| (3) |
14:30-15:00 |
Degrees of freedom in submodular regularization |
Kentaro Minami, Fumiyasu Komaki (The University of Tokyo) |
| |
15:00-15:15 |
Break ( 15 min. ) |
| (4) |
15:15-16:15 |
Invited talk
Yuichi Yoshida (National Institute of Informatics)
"Constant time algorithms in machine learning" |
| |
16:15-16:30 |
Break ( 15 min. ) |
Mon, Mar 6 PM Deep learning
Chair: Taiji Suzuki
16:30 - 17:30 |
| (5) |
16:30-17:00 |
New Lerning Algorythm of Neural Network using Integral Representation and Kernel Herding |
Takuo Matsubara, Sho Sonoda, Noboru Murata (Waseda Univ.) |
| (6) |
17:00-17:30 |
Recurrent Neural Networks for task-evoked fMRI data classification |
Koya Ohashi (Tokyo Tech), Taiji Suzuki (Tokyo Tech/JST/RIKEN) |
Tue, Mar 7 AM Optimization method
Chair: Shunichi Nomura
10:00 - 12:00 |
| (7) |
10:00-10:30 |
Safe Pruning Rule for Predictive Sequential Pattern Mining and Its Application to Bio-logging Data Analysis |
Kaoru Kishimoto (NITech), Masayuki Karasuyama (NIT), Kazuya Nakagawa (NITech), Kotaro Kimura (Osaka Univ.), Ken Yoda (Nagoya Univ.), Yuta Umezu, Shinsuke Kajioka (NITech), Koji Tsuda (UTokyo), Ichiro Takeuchi (NITech) |
| (8) |
10:30-11:00 |
Doubly Accelerated Stochastic Variance Reduced Gradient Method for Regularized Empirical Risk Minimization |
Tomoya Murata, Taiji Suzuki (Tokyo Tech) |
| (9) |
11:00-11:30 |
A study on minimizing size of sparse model optimization problem: exploiting safe rules for keeping and removing variables |
Masayuki Karasuyama (NIT/NIMS/JST), Atsushi Shibagaki (NIT), Ichiro Takeuchi (NIT/NIMS/RIKEN/NIMS) |
| (10) |
11:30-12:00 |
A stochastic optimization method and generalization bounds for voting classifiers by continuous density functions |
Atsushi Nitanda (Tokyo Tech./NTTDATA MSI), Taiji Suzuki (Tokyo Tech./JST/RIKEN) |
| |
12:00-13:00 |
Lunch break ( 60 min. ) |
Tue, Mar 7 PM Applications of machine learning
Chair: Taiji Suzuki
13:00 - 14:30 |
| (11) |
13:00-13:30 |
Classification of In-week and -day Patterns of Ambulatory Activity Using a Hierarchical Topic Model and Interpolation of Step Counts in Missing Days |
Shunichi Nomura (Tokyo Tech.), Michiko Watanabe, Yuko Oguma (Keio Univ.) |
| (12) |
13:30-14:00 |
Topic model for analysis of gene expression data |
Koji Iwayama (Ryukoku Univ.), Atsushi J. Nagano (Ryukoku Univ./Kyoto Univ.) |
| (13) |
14:00-14:30 |
CTC network with explicit representation vector of Markov property |
Yuta Kawachi, Taichi Asami, Yoshikazu Yamaguchi, Yushi Aono (NTT) |