Committee |
Date Time |
Place |
Paper Title / Authors |
Abstract |
Paper # |
PRMU, IBISML, IPSJ-CVIM |
2024-03-03 15:30 |
Hiroshima |
Hiroshima Univ. Higashi-Hiroshima campus (Primary: On-site, Secondary: Online) |
Randomized Channel-pass Filter for Explaining Black-box Models Daiki Nisawa, Hirotaka Hachiya (Wakayama Univ.) IBISML2023-42 |
(To be available after the conference date) [more] |
IBISML2023-42 pp.14-20 |
NC, IBISML, IPSJ-BIO, IPSJ-MPS [detail] |
2022-06-27 14:00 |
Okinawa |
(Primary: On-site, Secondary: Online) |
Transformer-Based Fully Trainable Model for Point Process with Past Sequence-Representative Vector Fumiya Nishizawa, Sujun Hong, Hirotaka Hachiya (Graduate School of System Engineering, Wakayama University) NC2022-1 IBISML2022-1 |
Recently, a Transformer-based partially trainable point process has been proposed, where a feature vector is extracted f... [more] |
NC2022-1 IBISML2022-1 pp.1-5 |
NC, IBISML, IPSJ-MPS, IPSJ-BIO [detail] |
2019-06-17 14:15 |
Okinawa |
Okinawa Institute of Science and Technology |
Triple GANs with adversarial disturbances for discriminative anomaly detection Hirotaka Hachiya (Wakayama Univ.) IBISML2019-4 |
Anomaly detection (AD) is an important machine learning task to detect outliers given only normal training data---applie... [more] |
IBISML2019-4 pp.21-26 |
NC, MBE (Joint) |
2019-03-04 13:25 |
Tokyo |
University of Electro Communications |
Approximation of inverse embedding function using neural network, and its application for the interpretation of document data distribution Yuya Mizobuchi, Hirotaka Hachiya (Wakayama Univ.) NC2018-54 |
Recently, analysing the relation between human and compnies based those related documents such as tweets and web pages.
... [more] |
NC2018-54 pp.61-66 |
IBISML |
2012-03-13 16:30 |
Tokyo |
The Institute of Statistical Mathematics |
Feature Selection via L1-Penalized Squared-Loss Mutual Information Wittawat Jitkrittum, Hirotaka Hachiya, Masashi Sugiyama (Tokyo Inst. of Tech.) IBISML2011-107 |
Feature selection is a technique to screen out less important features.
Many existing supervised feature selection alg... [more] |
IBISML2011-107 pp.139-146 |
IBISML |
2011-11-09 15:45 |
Nara |
Nara Womens Univ. |
Relative Density-Ratio Estimation for Robust Distribution Comparison Makoto Yamada (Tokyo Inst. of Tech.), Taiji Suzuki (Univ. of Tokyo), Takafumi Kanamori (Nagoya Univ.), Hirotaka Hachiya, Masashi Sugiyama (Tokyo Inst. of Tech.) IBISML2011-46 |
Divergence estimators based on direct approximation of density-ratios
without going through separate approximation of n... [more] |
IBISML2011-46 pp.25-32 |
IBISML |
2011-11-09 15:45 |
Nara |
Nara Womens Univ. |
Modified Newton Approach to Policy Search Hirotaka Hachiya (Tokyo Inst. of Tech.), Tetsuro Morimura (IBM Japan), Takaki Makino (Univ. of Tokyo), Masashi Sugiyama (Tokyo Inst. of Tech.) IBISML2011-54 |
The natural policy gradient method was shown to be a useful approach to policy search in reinforcement learning. However... [more] |
IBISML2011-54 pp.79-85 |
IBISML |
2011-11-10 15:45 |
Nara |
Nara Womens Univ. |
Computationally Efficient Multi-Label Classification by Least-Squares Probabilistic Classifier Hyunha Nam, Hirotaka Hachiya, Masashi Sugiyama (Tokyo Inst. of Tech.) IBISML2011-73 |
[more] |
IBISML2011-73 pp.213-216 |
IBISML |
2011-06-21 14:15 |
Tokyo |
Takeda Hall |
Analysis and Improvement of Policy Gradient Estimation Tingting Zhao, Hirotaka Hachiya, Gang Niu, Masashi Sugiyama (Tokyo Inst. of Tech.) IBISML2011-12 |
Policy gradient is a useful model-free reinforcement learning approach,
but it tends to suffer from instability of grad... [more] |
IBISML2011-12 pp.83-89 |
IBISML |
2011-03-29 15:50 |
Osaka |
Nakanoshima Center, Osaka Univ. |
Information-Maximization Clustering
-- Analytic Solution and Model Selection -- Masashi Sugiyama, Makoto Yamada, Manabu Kimura, Hirotaka Hachiya (Tokyo Inst. of Tech) IBISML2010-114 |
(Advance abstract in Japanese is available) [more] |
IBISML2010-114 pp.69-76 |
IBISML |
2010-11-05 15:30 |
Tokyo |
IIS, Univ. of Tokyo |
[Poster Presentation]
Return distribution estimation with dynamic programming Tetsuro Morimura (IBM Japan), Masashi Sugiyama (Tokyo Tech), Hisashi Kashima (Univ. of Tokyo), Hirotaka Hachiya (Tokyo Tech), Toshiyuki Tanaka (Kyoto Univ.) IBISML2010-98 |
(Advance abstract in Japanese is available) [more] |
IBISML2010-98 pp.283-290 |
IBISML |
2010-06-15 15:25 |
Tokyo |
Takeda Hall, Univ. Tokyo |
New Feature Selection Method for Reinforcement Learning
-- Conditional Mutual Information Reveals Implicit State-Reward Dependency -- Hirotaka Hachiya, Masashi Sugiyama (Tokyo Insst. of Tech.) IBISML2010-21 |
Model-free reinforcement learning (RL) is a machine learning approach to decision making in unknown environment. However... [more] |
IBISML2010-21 pp.139-146 |
NC, MBE (Joint) |
2009-03-12 14:50 |
Tokyo |
Tamagawa Univ. |
Adaptive Importance Sampling with Automatic Model Selection in Reward Weighted Regression Hirotaka Hachiya (Tokyo Inst. of Tech.), Jan Peters (Max Planck Inst. of Tech.), Masashi Sugiyama (Tokyo Inst. of Tech.) NC2008-145 |
Direct policy search is a useful framework of reinforcement learning in particular in continuous systems such as robot c... [more] |
NC2008-145 pp.249-254 |
NC, MBE (Joint) |
2009-03-12 15:40 |
Tokyo |
Tamagawa Univ. |
Statistical Active Learning for Efficient Value Function Approximation in Reinforcement Learning Takayuki Akiyama, Hirotaka Hachiya, Masashi Sugiyama (Tokyo Inst. of Tech.) NC2008-147 |
Appropriately designing sampling policies is highly important for obtaining better control policies in reinforcement lea... [more] |
NC2008-147 pp.261-266 |
MBE, NC (Joint) |
2007-12-22 15:45 |
Aichi |
|
Adaptive Importance Sampling with Automatic Model Selection in Value Function Approximation Hirotaka Hachiya, Takayuki Akiyama, Masashi Sugiyama (Tokyo Inst. of Tech.) NC2007-84 |
Off-policy reinforcement learning is aimed at efficiently reusing data samples gathered in the past. A common approach i... [more] |
NC2007-84 pp.75-80 |