Title:
EXTREME LEARNING MACHINE TRAINING METHOD, EXTREME LEARNING MACHINE TRAINING APPARATUS, AND TERMINAL DEVICE
Document Type and Number:
WIPO Patent Application WO/2022/155787
Kind Code:
A1
Abstract:
Disclosed are an extreme learning machine training method, an extreme learning machine training apparatus, and a terminal device. The method comprises: acquiring a target training sample set, which comprises at least two target sub-training sample sets; training each original extreme learning machine by using each of the target sub-training sample sets to obtain an original weight value corresponding to each target sub-training sample set; and fusing the original weight values to obtain a target fused weight value, and then constructing a target extreme learning machine according to the target fused weight value. According to the present application, the problems of a relatively long time and a relatively low efficiency of training a high-dimensional training sample insofar as the high accuracy of an existing extreme learning machine is ensured are solved.
Inventors:
HE YULIN (CN)
LI XU (CN)
LI XU (CN)
Application Number:
PCT/CN2021/072702
Publication Date:
July 28, 2022
Filing Date:
January 19, 2021
Export Citation:
Assignee:
UNIV SHENZHEN (CN)
International Classes:
G06Q10/06
Foreign References:
| CN108985574A | 2018-12-11 | |||
| CN110322933A | 2019-10-11 | |||
| US20110167024A1 | 2011-07-07 |
Attorney, Agent or Firm:
SHENZHEN ZHONGYI UNION INTELLECTUAL PROPERTY AGENCY CO., LTD. (CN)
Download PDF:
Previous Patent: PRECISE LINEAR FEEDBACK AMPLIFICATION CIRCUIT
Next Patent: AMBIENT FEATURE ACTIVE SENSING BASED CONTROL MODULE
Next Patent: AMBIENT FEATURE ACTIVE SENSING BASED CONTROL MODULE