Committee |
Date Time |
Place |
Paper Title / Authors |
Abstract |
Paper # |
NC |
2007-03-14 10:10 |
Tokyo |
Tamagawa University |
Unbiased Learning for Hierarchical Models Masashi Sekino, Katsumi Nitta (Tokyo Tech) |
It is known that overfitting occurs when a conventional statistical learning method such as maximum likelihood estimatio... [more] |
NC2006-136 pp.109-114 |
NC |
2007-03-14 10:40 |
Tokyo |
Tamagawa University |
Optimization of the support vector machine's kernels using a feedforward neural network Shin'ichi Tamura (DENSO CORP.) |
A novel approach to determining the optimal kernel parameters of Support Vector Machines is proposed. Based on the fact ... [more] |
NC2006-137 pp.115-120 |
NC |
2007-03-14 11:00 |
Tokyo |
Tamagawa University |
Split Parameter Optimization for Multiclass Support Vector Machines Keita Tsubota, Mauricio Kugler (NIT), Anto Satriyo Nuguroho (Chukyo Univ.), Susumu Kuroyanagi, Akira Iwata (NIT) |
A support vector machine (SVM) using Gaussian kernel function presents two parameters (C and sigma) that need to be tune... [more] |
NC2006-138 pp.121-126 |
NC |
2007-03-14 11:20 |
Tokyo |
Tamagawa University |
On Variable Selection in Decomposition Methods for Support Vector Machines
-- Proposal and Experimental Evaluation of a Novel Variable Selection based on Conjugate Gradient Method -- Yusuke Kawazoe (Kyushu Univ.), Masashi Kuranoshita (FUJIFILM), Norikazu Takahashi, Jun'ichi Takeuchi (Kyushu Univ.) |
Learning of a support vector machine (SVM) is formulated as a quadratic
programming (QP) problem. Decomposition method... [more] |
NC2006-139 pp.127-132 |
NC |
2007-03-14 11:40 |
Tokyo |
Tamagawa University |
Characteristics of Hopfield Model Using the Maximal Margin Learning Hitoshi Akimoto, Motonobu Hattori (University of Yamanashi) |
[more] |
NC2006-140 pp.133-138 |
NC |
2007-03-14 13:00 |
Tokyo |
Tamagawa University |
Effect of Sparse Encoding in the Layered Associative Network Kazuya Ishibashi (Univ. Tokyo), Kosuke Hamaguchi (RIKEN), Masato Okada (Univ. Tokyo/ RIKEN) |
A synfire chain is a simple neural network model which can generate stable synchronous firings called a pulse packet and... [more] |
NC2006-141 pp.139-144 |
NC |
2007-03-14 13:20 |
Tokyo |
Tamagawa University |
Chaotic Complex-valued Associative Memory Masao Nakada, Yuko Osana (Tokyo Tec. Univ.) |
[more] |
NC2006-142 pp.145-150 |
NC |
2007-03-14 13:40 |
Tokyo |
Tamagawa University |
Progressive contrastive divergence method Kazuya Takabatake, Shotaro Akaho (AIST) |
[more] |
NC2006-143 pp.151-154 |
NC |
2007-03-14 14:00 |
Tokyo |
Tamagawa University |
Analysis of Exchange Ratio for Heat Bath Type in Exchange Monte Carlo Method Kenji Nagata, Sumio Watanabe (Tokyo Inst. of Tech.) |
[more] |
NC2006-144 pp.155-160 |
NC |
2007-03-14 14:30 |
Tokyo |
Tamagawa University |
An ICA method of neural circuit for self-organization of hierarchical generative model Yuuji Ichisugi (AIST) |
An ICA method for BESOM model is presented.
This mechanism
self-organizes hierarchical generative model of the world. [more] |
NC2006-145 pp.161-166 |
NC |
2007-03-14 14:50 |
Tokyo |
Tamagawa University |
Independent component analysis for nonminimum phase systems using state space models Shuichi Fukunaga, Kenji Fujimoto (Nagoya Univ.) |
[more] |
NC2006-146 pp.167-171 |
NC |
2007-03-14 15:10 |
Tokyo |
Tamagawa University |
Markov and semi-Markov switching of source appearances for non-stationary independent component analysis Junichiro Hirayama (NAIST/JSPS), Shin-ichi Maeda, Shin Ishii (NAIST) |
Independent Component Analysis (ICA) is currently the most popularly used approach to blind source separation (BSS), the... [more] |
NC2006-147 pp.173-178 |
NC |
2007-03-14 15:30 |
Tokyo |
Tamagawa University |
A Study of Policy-Gradient Methods in Non-Markov Decision Porcesses
-- Application to a Curling Game -- Harukazu Igarashi (Shibaura Inst. Tech.), Seiji Ishihara (Kinki Univ.), Masaomi Kimura (Shibaura Inst. Tech.) |
There are two approaches to reinforcement learning: value-based methods and policy-gradient methods. Baird and Moore pro... [more] |
NC2006-148 pp.179-184 |
NC |
2007-03-14 16:00 |
Tokyo |
Tamagawa University |
Action-Oriented State Coding by Neighbourhood Component Analysis Makoto Otsuka (OIST/NAIST), Eiji Uchibe (OIST), Kenji Doya (OIST/NAIST) |
[more] |
NC2006-149 pp.185-190 |
NC |
2007-03-14 16:20 |
Tokyo |
Tamagawa University |
Prediction of the optimal parameter values in reinforcement learning based on interdependency Keiji Kamei, Masumi Ishikawa (KIT) |
Reinfoecement learning is that we have to specify parameters values without prior information. We proposed to optimize t... [more] |
NC2006-150 pp.191-196 |
NC |
2007-03-14 16:40 |
Tokyo |
Tamagawa University |
Learning of a robust controller for a biped robot based on sample reuse reinforcement learning method Tsuyoshi Ueno (NAIST), Yutaka Nakamura, Takashi Takuma (Osaka Univ.), Tomohiro Shibata (NAIST), Koh Hosoda (Osaka Univ.), Shin Ishii (NAIST) |
[more] |
NC2006-151 pp.197-202 |
NC |
2007-03-14 17:00 |
Tokyo |
Tamagawa University |
Acquiring vermicular motion of a Looper-like robot based on the CPG-Actor-Critic method Kenji Makino (NAIST), Yutaka Nakamura (Osaka Univ.), Tomohiro Shibata, Shin Ishii (NAIST) |
Adaptability to the environment is crucial for mobile robots, because the circumstance, including the body of the robot,... [more] |
NC2006-152 pp.203-208 |
NC |
2007-03-15 09:30 |
Tokyo |
Tamagawa University |
Optimality of reaching movements based on energetic cost under the influence of signal-dependent noise Yoshiaki Taniai, Jun Nishii (Yamaguchi Univ.) |
[more] |
NC2006-153 pp.1-5 |
NC |
2007-03-15 09:50 |
Tokyo |
Tamagawa University |
Selection of leg swing trajectorys based on metabolic cost Akihiro Fujii, Hiromichi Suenaga, Yoshiaki Taniai, Jun Nishii (Yamgaguchi Univ) |
When we walk, there are infinite number of possible trajectories for leg swing. How does the brain design the trajectory... [more] |
NC2006-154 pp.7-12 |
NC |
2007-03-15 10:10 |
Tokyo |
Tamagawa University |
An analysis of the kip motion on the horizontal bar based on the criterion considering energetic cost Miyo Takeuchi, Taiga Yamasaki, Kiyoshi Gotoh, Masahiro Kaneda, Xin Xin (Okayama Prefectural Univ.) |
[more] |
NC2006-155 pp.13-18 |