Title:
LEARNING-FINISHED MODEL INTEGRATION METHOD, DEVICE, PROGRAM, IC CHIP, AND SYSTEM
Document Type and Number:
WIPO Patent Application WO/2019/026703
Kind Code:
A1
Abstract:
[Problem] To create a single learning-finished model by integrating multiple learning-finished models obtained by using a new machine learning framework having a tree structure [Solution] The present invention provides a learning-finished model integration method for integrating a plurality of different learning-finished models obtained by causing a learning model having a tree structure to learn a predetermined data group, the learning-finished model integration method being provided with: a data reading step in which data on the plurality of different learning-finished models is read from a predetermined storage unit; and an integrating step in which, with respect to respective nodes constituting the tree structure related to the plurality of different learning-finished models, the plurality of different learning-finished models are integrated into a single learning-finished model by duplicating a node when the node is present in only one learning-finished mode, and by integrating nodes when the corresponding nodes are respectively present at the positions corresponding to the plurality of learning-finished models.
Inventors:
KIM CHYON HAE (JP)
IDESAWA JUNICHI (JP)
SUGAWARA SHIMON (JP)
IDESAWA JUNICHI (JP)
SUGAWARA SHIMON (JP)
Application Number:
PCT/JP2018/027774
Publication Date:
February 07, 2019
Filing Date:
July 24, 2018
Export Citation:
Assignee:
AISING LTD (JP)
International Classes:
G06N99/00
Foreign References:
| JP2016173686A | 2016-09-29 | |||
| JP2016173686A | 2016-09-29 |
Other References:
CHYON HAE KIM , IPSJ SIG TECHNICAL REPORT, MATHEMATICAL MODELING AND PROBLEM SOLVING (MPS, vol. 2017 -MP-112, no. 25, February 2017 (2017-02-01), pages 1 - 5, XP009515834, ISSN: 2188-8833
OTANI, NORIKO ET AL.: "Examination of partial solution in symbiotic evolution for decision tree generation", IEICE TECHNICAL REPORT, vol. 103, no. 725, 9 March 2004 (2004-03-09), pages 101 - 106, XP009515743, ISSN: 0913-5685
IDESAWA, JUNICHI ET AL.: "Japan first'' ''world first'' machine control AI, DeepBinaryTree:DBT, ''A machine learning algorithm that can be implemented on lightweight edge devices", LECTURE MATERIALS, 27 April 2018 (2018-04-27), pages 1 - 28, XP009515772, Retrieved from the Internet [retrieved on 20180906]
See also references of EP 3663992A4
OTANI, NORIKO ET AL.: "Examination of partial solution in symbiotic evolution for decision tree generation", IEICE TECHNICAL REPORT, vol. 103, no. 725, 9 March 2004 (2004-03-09), pages 101 - 106, XP009515743, ISSN: 0913-5685
IDESAWA, JUNICHI ET AL.: "Japan first'' ''world first'' machine control AI, DeepBinaryTree:DBT, ''A machine learning algorithm that can be implemented on lightweight edge devices", LECTURE MATERIALS, 27 April 2018 (2018-04-27), pages 1 - 28, XP009515772, Retrieved from the Internet
See also references of EP 3663992A4
Attorney, Agent or Firm:
IIZUKA Shin-ichi et al. (JP)
Download PDF: