Committee |
Date Time |
Place |
Paper Title / Authors |
Abstract |
Paper # |
SS |
2025-03-10 10:45 |
Kagoshima |
(Kagoshima) |
Investigating the generalization and robustness performance of CNN models when using RAdam Yusuke Yasuda, Kozo Okano, shinpei Ogata (SU), Shin Nakajima (NII) |
[more] |
|
SS |
2025-03-10 17:20 |
Kagoshima |
(Kagoshima) |
A Robustness Testing Method for Large Language Models Using Multiple-Choice Questions and an Analysis of the Temperature Effects Tatsuto Ota, Kozo Okano, Shinpei Ogata (Shinshu Univ), Shin Nakajima (NII) |
[more] |
|
KBSE |
2025-01-23 15:05 |
Osaka |
Osaka Kyoiku University(Tennoji) (Osaka, Online) (Primary: On-site, Secondary: Online) |
Quality Management of AI Systems equipped with LLMs
-- Inside the AIQM Guideline -- Shin Nakajima, Koichi Konishi, Yoshiki Seo, Yutaka Oiwa (AIST) KBSE2024-43 |
A series of Machine Learning Quality Management Guidelines have been published since 2020 as a basis for developing mach... [more] |
KBSE2024-43 pp.25-30 |
KBSE, SS, IPSJ-SE [detail] |
2024-07-25 14:00 |
Hokkaido |
(Hokkaido, Online) (Primary: On-site, Secondary: Online) |
Privacy protection of training datasets in CNN transfer learning models Takumi Katsuie, Kozo Okano, Shinpei Ogata (Shinshu Univ.), Shin Nakajima (NII) SS2024-1 KBSE2024-7 |
Transfer learning, one of the machine learning methods, has attracted attention as a technique that can create highly ac... [more] |
SS2024-1 KBSE2024-7 pp.1-6 |
KBSE, SS, IPSJ-SE [detail] |
2024-07-25 14:30 |
Hokkaido |
(Hokkaido, Online) (Primary: On-site, Secondary: Online) |
Debugging and testing of training datasets using internal activation states of CNN models Daichi Ofuchi, Takumi Katsuie, Kozo Okano, Shinpei Ogata (Shinshu Univ.), Shin Nakajima (NII) SS2024-2 KBSE2024-8 |
Recently, Artificial Intelligence (AI) and machine learning technologies have made great achievements in various fields.... [more] |
SS2024-2 KBSE2024-8 pp.7-12 |
SS, DC |
2023-10-12 10:25 |
Nagano |
(Nagano, Online) (Primary: On-site, Secondary: Online) |
Robustness trends of DP-SGD, a machine learning with differential privacy Takahiro Kanki, Shinpei Ogata, Kozo Okano (Sinshu Univ), Shin Nakajima (NII) SS2023-28 DC2023-34 |
Although machine learning has been successful in various fields, there is a problem that an adversary can extract traini... [more] |
SS2023-28 DC2023-34 pp.38-43 |
SS |
2023-03-15 13:45 |
Okinawa |
(Okinawa, Online) (Primary: On-site, Secondary: Online) |
Improvement of Encoding and Ablation Methods in Fault Localization by Ablation Takuma Ikeda, Kozo Okano, Shinpei Ogata (Shinshu Univ.), Shin Nakajima (NII) SS2022-67 |
Spectrum-based Fault Localization (SFL) is a technique to locate faults in source code using execution traces. A method ... [more] |
SS2022-67 pp.121-126 |
DC, SS |
2022-10-25 13:00 |
Fukushima |
(Fukushima, Online) (Primary: On-site, Secondary: Online) |
[Invited Talk]
Privacy as Ethical Quality Characteristics of AI Shin Nakajima (NII) SS2022-25 DC2022-31 |
[more] |
SS2022-25 DC2022-31 p.22 |
DC, SS |
2022-10-25 14:15 |
Fukushima |
(Fukushima, Online) (Primary: On-site, Secondary: Online) |
Relationship between the Defects in Learning Programs and the Model Distortion on the Convolutional Neural Networks Takumi Tsuchiya, Kozo Okano, Shinpei Ogata (Shinshu Univ.), Shin Nakajima (NII) SS2022-26 DC2022-32 |
In recent years, the quality issue of machine learning software has become an important concern. When considering the qu... [more] |
SS2022-26 DC2022-32 pp.23-28 |
SS |
2022-03-07 11:20 |
Online |
Online (Online) |
Trace Ablation and Fault Localization per Method Using Machine Learning Models for Automatic Classification of Test Execution Results Takuma Ikeda, Kozo Okano, Shinpei Ogata (Shinshu Univ.), Shin Nakajima (NII) SS2021-44 |
The problem to solve automatically classifying the results of test executions is called the test oracle problem. This is... [more] |
SS2021-44 pp.13-18 |
SS, MSS |
2022-01-12 09:15 |
Nagasaki |
Nagasakiken-Kensetsu-Sogo-Kaikan Bldg. (Nagasaki, Online) (Primary: On-site, Secondary: Online) |
Execution-trace embedding using word-proximity metric for a method to automatically classify test results Takuma Ikeda, Kozo Okano, Shinpei Ogata (Shinshu Univ.), Shin Nakajima (NII) MSS2021-46 SS2021-33 |
The problem to solve automatically classifying the results of test executions is called the test oracle problem. This is... [more] |
MSS2021-46 SS2021-33 pp.83-88 |
SS |
2020-03-04 14:20 |
Okinawa |
(Okinawa) (Cancelled but technical report was issued) |
Distortion Metrics for Trained Machine Learning Models Shin Nakajima (NII) SS2019-44 |
[more] |
SS2019-44 pp.19-24 |
SS, MSS |
2020-01-15 10:30 |
Hiroshima |
(Hiroshima) |
MSS2019-51 SS2019-35 |
[more] |
MSS2019-51 SS2019-35 pp.61-66 |
SS |
2019-03-05 15:00 |
Okinawa |
(Okinawa) |
Quality Evaluation Asssurance Levels for Machine Learning Software Shin Nakajima (NII), Yoshiki Seo, Yutaka Oiwa, Yoshinao Isobe (AIST) SS2018-79 |
[more] |
SS2018-79 pp.163-168 |
KBSE, SC |
2018-11-09 13:00 |
Hyogo |
(Hyogo) |
Quality of Machine Learning Software: Products, Services, Platform Shin Nakajima (NII) KBSE2018-31 SC2018-26 |
[more] |
KBSE2018-31 SC2018-26 pp.19-24 |
SS |
2018-03-07 15:30 |
Okinawa |
(Okinawa) |
Uncertainty in Machine Learning Software from Quality Assurance Viewpoints Shin Nakajima (NII) SS2017-77 |
[more] |
SS2017-77 pp.75-80 |
SS |
2017-03-09 10:20 |
Okinawa |
(Okinawa) |
Debugging Pure Strategy Games Shin Nakajima (NII) SS2016-62 |
[more] |
SS2016-62 pp.13-18 |
KBSE, SS, IPSJ-SE [detail] |
2016-07-13 09:50 |
Hokkaido |
(Hokkaido) |
Testing Quasi-testable Core of Non-testable Program Shin Nakajima (NII), Bui Ngoc Hai (VNU) SS2016-1 KBSE2016-7 |
[more] |
SS2016-1 KBSE2016-7 pp.1-6 |
KBSE, SS, IPSJ-SE [detail] |
2015-07-24 10:40 |
Hokkaido |
(Hokkaido) |
Two-Staged Modeling Method with Alloy and Event-B Shin Nakajima (NII) SS2015-32 KBSE2015-25 |
[more] |
SS2015-32 KBSE2015-25 pp.161-166 |
SS |
2015-03-09 10:20 |
Okinawa |
OKINAWAKEN SEINENKAIKAN (Okinawa) |
SNIPER: An LLVM-based Automatic Fault Localization Tool for Imperative Programs Si-Mohamed Lamraoui, Shin Nakajima (NII) SS2014-57 |
[more] |
SS2014-57 pp.13-18 |