Committee 
Date Time 
Place 
Paper Title / Authors 
Abstract 
Paper # 
IBISML 
20210302 10:25 
Online 
Online 
Selective Inference for Convex Clustering Using Parametric Programming Yumehiro Omori, Yu Inatsu (Nitech), Ichiro Takeuchi (Nitech/RIKEN) IBISML202035 
Traditional statistical inference assumes that the hypothesis is predetermined and cannot be used as is for statistical ... [more] 
IBISML202035 pp.915 
IBISML 
20131112 15:45 
Tokyo 
Tokyo Institute of Technology, KuramaeKaikan 
[Poster Presentation]
NonConvex Optimization of Robust Support Vector Regression by Utilizing Parametric Programming Shinya Suzumura, Ichiro Takeuchi (Nagoya Inst. of Tech.), Masashi Sugiyama (Tokyo Inst. of Tech.) IBISML201345 
In this paper, we propose a novel optimization method for robust support vector regression(SVR) that has robustness to o... [more] 
IBISML201345 pp.6975 
IBISML 
20131113 15:45 
Tokyo 
Tokyo Institute of Technology, KuramaeKaikan 
[Poster Presentation]
Learning Common Features of Parametrized Tasks Ichiro Takeuchi, Tatsuya Hongo (Nagoya Inst. of Tech.), Masashi Sugiyama (Tokyo Inst. of Tech.), Shinichi Nakajima (Nikon) IBISML201366 
We introduce a novel formulation of multitask learning (MTL) called parametric task learning (PTL) that can systematica... [more] 
IBISML201366 pp.225232 
COMP 
20130624 15:10 
Nara 
Nara Women's University 
On a parametic integer programming algorithm using dualization of monotone Boolean function Norie Fu (NII), Takafumi Shibuta (Kyushu Univ.) COMP201323 
The {it parametric integer programing} is the integer programming with a vector of parameters as its righthandside vec... [more] 
COMP201323 pp.2330 
IBISML 
20121108 15:00 
Tokyo 
Bunkyo School Building, Tokyo Campus, Tsukuba Univ. 
Path following approach for efficient reweighted L1 minimization Yuki Shinmura, Ichiro Takeuchi (NIT) IBISML201271 
The problem of recovering sparse signals is an important topic in machine learning and compressed sensing literatures.
... [more] 
IBISML201271 pp.265270 
IBISML 
20121108 15:00 
Tokyo 
Bunkyo School Building, Tokyo Campus, Tsukuba Univ. 
MultiInstance SVM using Parametric Programming Naoki Ishihara, Saori Kurumi, Ichiro Takeuchi (NIT) IBISML201272 
In this paper, we propose a new optimization algorithm for multiple instance learning (MIL). MIL is a supervised learnin... [more] 
IBISML201272 pp.271276 
IBISML 
20121108 15:00 
Tokyo 
Bunkyo School Building, Tokyo Campus, Tsukuba Univ. 
Efficient SVM Bootstrap Computation by Parametric Programming Yoshiki Suzuki, Kohei Ogawa, Ichiro Takeuchi (Nagoya Inst. of Tech.) IBISML201273 
In this paper, we study statistical variability estimation of the support vector machine (SVM) by using bootstrap method... [more] 
IBISML201273 pp.277282 
IBISML 
20120619 10:00 
Kyoto 
Campus plaza Kyoto 
A study on an optimization algorithm for semisupervised SVM using parametric programing Kohei Ogawa, Ichiro Takeuchi (NIT), Masashi Sugiyama (Tokyo Tech) IBISML20121 
The goal of semisupervised learning is to incorporate unlabeled instances as well as labeled ones for improving classif... [more] 
IBISML20121 pp.18 
IBISML 
20111110 15:45 
Nara 
Nara Womens Univ. 
A Parametric Programming Approach for Outlier Detection and Robust Learning for Classification and Regression Ichiro Takeuchi (NIT) IBISML201181 
We study outlier detection and robust learning problem for support vector machine (SVM). In the literature there are two... [more] 
IBISML201181 pp.263269 
PRMU, IBISML, IPSJCVIM [detail] 
20110905 13:30 
Hokkaido 

[Invited Talk]
Optimal Solution Path Following Algorithm for Pattern Recognition and Machine Learning Ichiro Takeuchi (NIT) PRMU201166 IBISML201125 
Many pattern classification and machine learning algorithms are formulated as mathematical optimization problems. These ... [more] 
PRMU201166 IBISML201125 pp.5960 
IBISML 
20101104 15:00 
Tokyo 
IIS, Univ. of Tokyo 
[Poster Presentation]
A Study on Simultaneous Feature Selection for CostSensitive Classifiers Using MixedNorm Regularization Toru Sugiura, Kazuaki Koide, Tatsuya Hongo, Masayuki Karasuyama, Ichiro Takeuchi (NIT) IBISML201070 
Costsensitive learning is useful for binary classification when the
costs of missclassifications are not symmetric. I... [more] 
IBISML201070 pp.8390 
IBISML 
20101105 15:30 
Tokyo 
IIS, Univ. of Tokyo 
[Poster Presentation]
A Study on the Suboptimal Solution Path Algorithm Masayuki Karasuyama, Ichiro Takeuchi (NIT) IBISML201089 
The solution path algorithms in machine learning trace the exact optimal solutions by exploiting the piecewise linearity... [more] 
IBISML201089 pp.221230 
IBISML, PRMU, IPSJCVIM [detail] 
20100905 17:30 
Fukuoka 
Fukuoka Univ. 
A Study on Feature Selection Path for HighDimensional Local Classifiers Ichiro Takeuchi (NIT) PRMU201071 IBISML201043 
We study feature selection and weighting problems for localbased classifier. The proposed algorithm is formulated as a ... [more] 
PRMU201071 IBISML201043 pp.105112 