Committee |
Date Time |
Place |
Paper Title / Authors |
Abstract |
Paper # |
PRMU, IBISML, IPSJ-CVIM |
2024-03-04 10:40 |
Hiroshima |
Hiroshima Univ. Higashi-Hiroshima campus (Primary: On-site, Secondary: Online) |
Poisoning Attack on Fairness of Fair Classification Algorithm through Threshold Control Dai Shengtian, Akimoto Youhei (Univ. of Tsukuba/RIKEN), Jun Sakuma (Tokyo Tech./RIKEN), Fukuchi Kazuto (Univ. of Tsukuba/RIKEN) IBISML2023-47 |
The ethical issues of artificial intelligence have become more severe as machine learning is widely used in several fiel... [more] |
IBISML2023-47 pp.49-56 |
IBISML |
2023-12-21 11:20 |
Tokyo |
National Institute of Informatics (Primary: On-site, Secondary: Online) |
Classification Error Analysis under Covariate Shift between Non-absolutely Continuous Distributions through neighbor-transfer-exponent Mitsuhiro Fujikawa, Youhei Akimoto (Univ. of Tsukuba), Jun Sakuma (Tokyo Inst. of Tech.), Kazuto Fukuchi (Univ. of Tsukuba) IBISML2023-39 |
Transfer learning is considered successful when increasing the source sample size decreases the target sample size neede... [more] |
IBISML2023-39 pp.58-65 |
IBISML |
2022-12-23 10:00 |
Kyoto |
Kyoto University (Primary: On-site, Secondary: Online) |
Zero-shot domain adaptation based on dual-level mix and contrast Yu Zhe, Fukuchi Kazuto, Sakuma Jun (Tsukuba Univ/Riken AIP) IBISML2022-53 |
[more] |
IBISML2022-53 pp.70-77 |
IBISML |
2022-12-23 10:30 |
Kyoto |
Kyoto University (Primary: On-site, Secondary: Online) |
Statistically Significant Concept-based Explanation via Model Knockoffs Kaiwen Xu, Kazuto Fukuchi, Youhei Akimoto, Jun Sakuma (Univ. Tsukuba/RIKEN AIP) IBISML2022-54 |
[more] |
IBISML2022-54 pp.78-85 |
IBISML |
2022-12-23 10:50 |
Kyoto |
Kyoto University (Primary: On-site, Secondary: Online) |
Interpretable Deep Image Classifier with Class-distinguishable Concept Text Kazuhiro Saito, Kazuto Fukuchi (Univ.Tsukuba), Jun Sakuma (Univ.Tsukuba/RIKEN) IBISML2022-55 |
(To be available after the conference date) [more] |
IBISML2022-55 pp.86-93 |
IBISML |
2022-01-17 11:20 |
Online |
Online |
CAMRI Loss: Class-wise Additive Angular Margin Loss for Improving Recall of a Specific Class Daiki Nishiyama (Univ. Tsukuba), Fukuchi Kazuto, Yohei Akimoto, Jun Sakuma (Univ. Tsukuba/RIKEN) IBISML2021-22 |
In real-world applications of multiclass classification models, there is a need to increase the recall of classes where ... [more] |
IBISML2021-22 pp.29-36 |
IBISML |
2022-01-18 13:20 |
Online |
Online |
IBISML2021-24 |
We aim to explain a black-box classifier with the form: `data X is classified as class Y because X has A, B and does not... [more] |
IBISML2021-24 pp.45-53 |
IBISML |
2020-10-22 14:50 |
Online |
Online |
Suppressing explanations with irrelevant concepts in deep learning Munemasa Tomohiro (Tsukuba Univ), Fukuchi Kazuto, Akimoto Yohei, Sakuma Jun (Tsukuba Univ/Riken AIP) IBISML2020-32 |
TCAV [1], which is an explanation method using a concept that humans easily understand for deep learning models, concept... [more] |
IBISML2020-32 pp.61-68 |
IBISML |
2017-11-09 13:00 |
Tokyo |
Univ. of Tokyo |
Online Optimization Method for Generalized $ell_1$ Regularized Problems Yoshihiro Nakazato, Kazuto Fukuchi (Tsukuba Univ.), Jun Sakuma (Tsukuba Univ./Riken/JST) IBISML2017-47 |
Structured sparse regularization is vital to enhance the precision and the interpretability of the model by introducing ... [more] |
IBISML2017-47 pp.93-100 |
IBISML |
2017-11-09 13:00 |
Tokyo |
Univ. of Tokyo |
Consequently Fair Contextual Bandit Learning Kazuto Fukuchi (Univ. of Tsukuba), Jun Sakuma (Univ. of Tsukuba/JST/RIKEN) IBISML2017-53 |
Fairness in machine learning is being recognized as an important field. It requires that the consequent decisions made b... [more] |
IBISML2017-53 pp.139-146 |
IBISML |
2016-11-16 15:00 |
Kyoto |
Kyoto Univ. |
Proximal Average Accelerated Proximal Gradient Algorithm with Adaptive Restart Yoshihiro Nakazato, Kazuto Fukuchi, Jun Sakuma (Univ. Tsukuba) IBISML2016-55 |
When using multiple regularizers, their proximal mapping is not easily available in closed form.
The method to calculat... [more] |
IBISML2016-55 pp.65-71 |
IBISML |
2016-11-17 14:00 |
Kyoto |
Kyoto Univ. |
Minimax optimal estimator for additively decomposable scalar functionals of discrete distributions Kazuto Fukuchi, Jun Sakuma (Univ. of Tsukuba) IBISML2016-82 |
We deal with a problem of estimating additively decomposable scalar functionals from a set of $n$ iid samples drawn from... [more] |
IBISML2016-82 pp.259-265 |
IBISML |
2014-11-17 17:00 |
Aichi |
Nagoya Univ. |
[Poster Presentation]
Differential Privacy on Linear Regression Model of Crowdsensing Tran Quang Khai, Kazuto Fukuchi, Jun Sakuma (Univ. of Tsukuba) IBISML2014-47 |
Learning statistic models using the data collected from crowd is one of the important tasks in the crowdsensing. Crowdse... [more] |
IBISML2014-47 pp.95-102 |
PRMU, IBISML, IPSJ-CVIM [detail] |
2014-09-01 17:30 |
Ibaraki |
|
Neutralized Empirical Risk Minimization with Covariance-based Neutrality Risk Kazuto Fukuchi, Jun Sakuma (Univ. of Tsukuba) PRMU2014-48 IBISML2014-29 |
In order to apply machine learning algorithms to real world problems, it is necessary to ensure that discrimination, unf... [more] |
PRMU2014-48 IBISML2014-29 pp.93-100 |