Title:
LEARNING PROCESSING DEVICE AND LEARNING PROCESSING METHOD FOR POOLING HIERARCHICALLY STRUCTURED GRAPH DATA ON BASIS OF GROUPING MATRIX, AND METHOD FOR TRAINING ARTIFICIAL INTELLIGENCE MODEL
Document Type and Number:
WIPO Patent Application WO/2023/224307
Kind Code:
A1
Abstract:
Disclosed are a learning processing device and method for pooling hierarchically structured graph data on the basis of a grouping matrix, and a method for training an artificial intelligence model. A learning processing device, according to one embodiment of the present invention, comprises: a memory; and a processor that communicates with the memory, wherein the processor generates a grouping matrix in a secondary form performing grouping on the basis of the similarity of pairwise nodes as graph data is inputted into a first artificial intelligence model which has been pre-trained, and decomposes the grouping matrix to generate a pooling matrix, wherein a pooling operator is obtained by decomposing the grouping matrix into a square root form.
Inventors:
KO SUNG MOON (KR)
CHO SUNGJUN (KR)
JEONG DAEWOONG (KR)
HAN SEHUI (KR)
LEE MOONTAE (KR)
LEE HONGLAK (KR)
CHO SUNGJUN (KR)
JEONG DAEWOONG (KR)
HAN SEHUI (KR)
LEE MOONTAE (KR)
LEE HONGLAK (KR)
Application Number:
PCT/KR2023/006356
Publication Date:
November 23, 2023
Filing Date:
May 10, 2023
Export Citation:
Assignee:
LG MAN DEVELOPMENT INSTITUTE CO LTD (KR)
International Classes:
G06N3/08; G06F17/16; G06N3/04
Other References:
BIANCHI FILIPPO MARIA, GRATTAROLA DANIELE, ALIPPI CESARE: "Spectral Clustering with Graph Neural Networks for Graph Pooling", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, ARXIV.ORG, ITHACA, 29 December 2020 (2020-12-29), Ithaca, XP093109773, DOI: 10.48550/arxiv.1907.00481
SONG YUE; SEBE NICU; WANG WEI: "Why Approximate Matrix Square Root Outperforms Accurate SVD in Global Covariance Pooling?", 2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), IEEE, 10 October 2021 (2021-10-10), pages 1095 - 1103, XP034093885, DOI: 10.1109/ICCV48922.2021.00115
JIN WEI ET AL.: "Node Similarity Preserving Graph Convolutional Networks", PROCEEDINGS OF THE 1ST WORKSHOP ON MACHINE LEARNING AND SYSTEMS, ACMPUB27, NEW YORK, NY, USA, 8 March 2021 (2021-03-08) - 26 April 2021 (2021-04-26), New York, NY, USA , pages 148 - 156, XP058565627, ISBN: 978-1-4503-8298-4, DOI: 10.1145/3437963.3441735
LI PEIHUA; XIE JIANGTAO; WANG QILONG; GAO ZILIN: "Towards Faster Training of Global Covariance Pooling Networks by Iterative Matrix Square Root Normalization", 2018 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, IEEE, 18 June 2018 (2018-06-18), pages 947 - 955, XP033476056, DOI: 10.1109/CVPR.2018.00105
DAVIDE BACCIU; LUIGI DI SOTTO: "A Non-Negative Factorization approach to node pooling in Graph Convolutional Neural Networks", ARXIV.ORG, 7 September 2019 (2019-09-07), XP081474944
SONG YUE; SEBE NICU; WANG WEI: "Why Approximate Matrix Square Root Outperforms Accurate SVD in Global Covariance Pooling?", 2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), IEEE, 10 October 2021 (2021-10-10), pages 1095 - 1103, XP034093885, DOI: 10.1109/ICCV48922.2021.00115
JIN WEI ET AL.: "Node Similarity Preserving Graph Convolutional Networks", PROCEEDINGS OF THE 1ST WORKSHOP ON MACHINE LEARNING AND SYSTEMS, ACMPUB27, NEW YORK, NY, USA, 8 March 2021 (2021-03-08) - 26 April 2021 (2021-04-26), New York, NY, USA , pages 148 - 156, XP058565627, ISBN: 978-1-4503-8298-4, DOI: 10.1145/3437963.3441735
LI PEIHUA; XIE JIANGTAO; WANG QILONG; GAO ZILIN: "Towards Faster Training of Global Covariance Pooling Networks by Iterative Matrix Square Root Normalization", 2018 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, IEEE, 18 June 2018 (2018-06-18), pages 947 - 955, XP033476056, DOI: 10.1109/CVPR.2018.00105
DAVIDE BACCIU; LUIGI DI SOTTO: "A Non-Negative Factorization approach to node pooling in Graph Convolutional Neural Networks", ARXIV.ORG, 7 September 2019 (2019-09-07), XP081474944
Attorney, Agent or Firm:
KIM, Myung Hoon (KR)
Download PDF:
Previous Patent: METHOD AND DEVICE FOR PERFORMING SENSING PROCEDURE IN WIRELESS LAN SYSTEM
Next Patent: NASAL BONE IMPLANT
Next Patent: NASAL BONE IMPLANT