Title:
PREDICTION MODEL TRAINING METHOD, DATA MONITORING METHOD, APPARATUSES, DEVICE AND MEDIUM
Document Type and Number:
WIPO Patent Application WO/2019/056499
Kind Code:
A1
Abstract:
A prediction model training method, a data monitoring method, apparatuses, a device and a medium. The prediction model training method comprises: performing time stamping on original service data and dividing same according to a preset period to obtain training service data carrying a timing status (S110); dividing the training service data into a training set and a test set according to a preset ratio (S120); using a long-short-time memory cyclic neural network model to train the training set to obtain an original prediction model (S130); and using the test set to test the original prediction model to obtain a target prediction model (S140). The prediction model training method has the advantages of strong timing performance and high accuracy when performing prediction.
More Like This:
Inventors:
ZHANG CHUAN (CN)
GU QINGSHAN (CN)
JIN XIN (CN)
WEN SHAN AN (CN)
LI LEI (CN)
WANG DINGXIN (CN)
LI YONGZHANG (CN)
GU QINGSHAN (CN)
JIN XIN (CN)
WEN SHAN AN (CN)
LI LEI (CN)
WANG DINGXIN (CN)
LI YONGZHANG (CN)
Application Number:
PCT/CN2017/108533
Publication Date:
March 28, 2019
Filing Date:
October 31, 2017
Export Citation:
Assignee:
PING AN TECH SHENZHEN CO LTD (CN)
International Classes:
G06Q10/06
Foreign References:
CN106886846A | 2017-06-23 | |||
CN105786830A | 2016-07-20 | |||
CN106603293A | 2017-04-26 | |||
US20150088606A1 | 2015-03-26 |
Attorney, Agent or Firm:
ZHONGDING INTELLECTUAL PROPERTY AGENCY (CN)
Download PDF: