Title:
COMPUTER SYSTEM AND MODEL TRAINING METHOD
Document Type and Number:
WIPO Patent Application WO/2024/062639
Kind Code:
A1
Abstract:
This system manages a first model for solving one or more tasks, and a second model for generating replay input data obtained by reproducing input data constituting training data used for learning of past tasks. When receiving new training data for a new task, the system uses the first model and the second model to generate replay training data, executes training processing for updating the first model by using the new training data and the replay training data, calculates an indicator representing uncertainty of the replay input data on the basis of an output obtained by inputting the replay input data into the updated first model, selects replay training data to be used for training on the basis of the indicator, and executes training processing by using the new training data and the selected replay training data.
More Like This:
| WO/2024/059658 | TRANSOMIC SYSTEMS AND METHODS OF THEIR USE |
| JP2024005890 | Learning device and learning method |
| WO/2017/070427 | AUTOMATIC PREDICTION OF ACOUSTIC ATTRIBUTES FROM AN AUDIO SIGNAL |
Inventors:
OGINO MASAHIRO (JP)
LI ZISHENG (JP)
LI ZISHENG (JP)
Application Number:
PCT/JP2023/001800
Publication Date:
March 28, 2024
Filing Date:
January 20, 2023
Export Citation:
Assignee:
HITACHI LTD (JP)
International Classes:
G06N20/00
Domestic Patent References:
| WO2017145960A1 | 2017-08-31 |
Other References:
SHIN HANUL, JUNG KWON LEE, KIM JAEHONG, KIM JIWON: "Continual Learning with Deep Generative Replay", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, ARXIV.ORG, ITHACA, 12 December 2017 (2017-12-12), Ithaca, XP093143138, Retrieved from the Internet [retrieved on 20240319], DOI: 10.48550/arxiv.1705.08690
KAWASHIMA HIRONO, KAWANO MAKOTO, NAKAZAWA JIN: "Impact of Replay Method for Class Difficulty on Incremental Scenarios in Class Incremental Learning of Continual Learning", THE 36TH ANNUAL CONFERENCE OF THE JAPANESE SOCIETY FOR ARTIFICIAL INTELLIGENCE, 14 June 2022 (2022-06-14), pages 1 - 4, XP093150070
KAWASHIMA HIRONO, KAWANO MAKOTO, NAKAZAWA JIN: "Impact of Replay Method for Class Difficulty on Incremental Scenarios in Class Incremental Learning of Continual Learning", THE 36TH ANNUAL CONFERENCE OF THE JAPANESE SOCIETY FOR ARTIFICIAL INTELLIGENCE, 14 June 2022 (2022-06-14), pages 1 - 4, XP093150070
Attorney, Agent or Firm:
TOU-OU PATENT FIRM (JP)
Download PDF:
Previous Patent: AUTONOMOUS TRAVEL-TYPE CLEANER
Next Patent: HOUSEKEEPING ASSISTANCE SYSTEM AND PROGRAM
Next Patent: HOUSEKEEPING ASSISTANCE SYSTEM AND PROGRAM