Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
METHOD FOR OPTIMIZING HYPER PARAMETER OF LIGHTWEIGHT ARTIFICIAL INTELLIGENCE ALGORITHM BY USING GENETIC ALGORITHM
Document Type and Number:
WIPO Patent Application WO/2022/124449
Kind Code:
A1
Abstract:
A method for optimizing a hyper parameter of a lightweight artificial intelligence algorithm by using a genetic algorithm is provided. A method for optimizing an artificial intelligence algorithm, according to an embodiment of the present invention, generates an entity population in which parameters of the artificial intelligence algorithm are defined as entities, evolves the entities by applying a genetic algorithm to the entity population, and sets parameters of the artificial intelligence algorithm by selecting one of the evolved entities. Therefore, a hyper parameter used in a lightweight artificial intelligence algorithm can be optimized using the genetic algorithm, and thus the performance of a lightweight artificial intelligence algorithm model is improved.

Inventors:
KIM BYUNG SOO (KR)
HWANG TAE HO (KR)
JANG YOUNG JONG (KR)
LEE JEA HACK (KR)
JEON SEOK HUN (KR)
Application Number:
PCT/KR2020/018130
Publication Date:
June 16, 2022
Filing Date:
December 11, 2020
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
KOREA ELECTRONICS TECHNOLOGY (KR)
International Classes:
G06N3/12; G06N3/08
Foreign References:
US20150106310A12015-04-16
US20130185234A12013-07-18
US8756179B22014-06-17
US20050246297A12005-11-03
KR20170072110A2017-06-26
Attorney, Agent or Firm:
NAM, Choong Woo (KR)
Download PDF: