インフルエンザ予測モデルの生成方法、装置及びコンピュータ可読記憶媒体

Japanese Patent JP6815708

B2

The present application discloses a method for generating an influenza prediction model. The method comprises: acquiring data concerning the percentage of influenza-like cases within a plurality of successive time units, and establishing an autoregressive integrated moving average (ARIMA) model; acquiring, according to a public opinion keyword, a public opinion data sequence within the plurality of time units, using public opinion data in the public opinion data sequence as a prediction feature, to train an extreme gradient boosting (xgboost) prediction model constructed on the basis of an xgboost algorithm, to determine model parameters; constructing, according to the ARIMA model and the xgboost prediction model, an influenza prediction model based on a Kalman filtering algorithm; and in the process of performing influenza prediction using the influenza prediction model, using a first prediction value of the ARIMA model as a measurement value of a state variable, and using a second prediction value of the xgboost prediction model as a priori estimation value of the state variable, to dynamically update the Kalman gain of the influenza prediction model. The present application further proposes an apparatus for generating an influenza prediction model and a computer readable storage medium. The present application improves the prediction accuracy of an influenza prediction model.

Lee Ching
Xu Liang
阮 暁▲うぇん▼
Xiao Kyo

JP2019556833A

January 20, 2021

August 24, 2018

Click for automatic bibliography generation  

PING AN TECHNOLOGY (SHENZHEN) CO.,LTD.

G16H50/80; G06F17/18

WO2019037260A1
KR1020140034994A

GINSBERG Jeremy,Detecting influenza epidemics using search engine query data,nature,2009年 2月19日,Vol.457,pp.1012-1015
荒牧 英治,ソーシャルメディアでインフルエンザ流行を捉える,岩波データサイエンス,株式会社岩波書店,2016年10月14日,Vol.4,pp.96-109

try international patent corporation

Previous Patent: 顔姿勢検出方法、装置及び記憶媒体

Next Patent: 使い捨て便座シート