Title:
FM MODEL BASED METHOD AND APPARATUS FOR PREDICTING MEDICAL HOT SPOT, AND COMPUTER DEVICE
Document Type and Number:
WIPO Patent Application WO/2021/139271
Kind Code:
A1
Abstract:
An FM model based method and apparatus for predicting a medical hot spot, and a computer device, relating to the field of artificial intelligence. In the method, an FM model is firstly applied to the prediction field of medical research hot spots in the medical field, which is suitable for processing sparse features, and a combination relationship between features can be excavated. In addition, compared with knowledge map prediction and SVD algorithm prediction, the FM model based method for predicting a medical hot spot can increase the construction features, and the increase in the features facilitates the achievement of a better result of the model. In the method, popular research relationships in the medical field can be calculated, possible future research hot spots can be predicted on the basis of the FM model, and an entity which is most likely to be researched in the future for a certain disease can be predicted. The present invention facilitates the searching of current research content by doctors and also helps to provide information of potential research hot spots to the doctors, and can further be applied to the field of blockchains, e.g. storing a trained model in a blockchain network.
Inventors:
CAO LIYU (CN)
Application Number:
PCT/CN2020/118914
Publication Date:
July 15, 2021
Filing Date:
September 29, 2020
Export Citation:
Assignee:
PING AN TECH SHENZHEN CO LTD (CN)
International Classes:
G06F16/36; G06F16/335; G06F16/9535; G06F40/295
Foreign References:
| CN111047406A | 2020-04-21 | |||
| CN111191136A | 2020-05-22 | |||
| CN110413946A | 2019-11-05 | |||
| CN109670054A | 2019-04-23 | |||
| US20200005196A1 | 2020-01-02 | |||
| US20170085929A1 | 2017-03-23 |
Attorney, Agent or Firm:
SHENZHEN MINGRIJINDIAN INTELLECTUAL PROPERTY AGENCY FIRM (GENERAL) (CN)
Download PDF: