Regional distribution of non-human H7N9 avian influenza virus detections in China and construction of a predictive model

Zeying Huang; Haijun Li; Beixun Huang

Otwarty dostęp

Regional distribution of non-human H7N9 avian influenza virus detections in China and construction of a predictive model

Zeying Huang

Haijun Li

oraz

Beixun Huang

| 05 lip 2021

Journal of Veterinary Research

Tom 65 (2021): Zeszyt 3 (September 2021)

O artykule

Poprzedni artykuł

Następny artykuł

Zacytuj

Udostępnij

Data publikacji: 05 lip 2021

Zakres stron: 253 - 264

Otrzymano: 04 sty 2021

Przyjęty: 10 cze 2021

DOI: https://doi.org/10.2478/jvetres-2021-0034

Słowa kluczowe
epidemiological survey, aetiological virus detection monitoring, wavelet packet decomposition, LS-SVM-ARIMA model, PSO algorithm

© 2021 Z. Huang et al. published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Regional distribution of H7N9 avian influenza virus detections in China

Regional distribution of H7N9 avian influenza virus in China from 2013 to 2020

Regional distribution of cumulative H7N9 avian influenza virus detections in China from January to December between 2013 and 2020

Regional distribution of H7N9 avian influenza virus detections before and after comprehensive H7N9 immunisation of at-risk animals in China

Temporal distribution of positive rate of animal H7N9 avian influenza virus detections in China from 2013 to 2020

Steps for constructing the LS-SVM-ARIMA combined model based on WPDLS-SVM – least squares support-vector machines; ARIMA – autoregressive integrated moving average

Low-frequency trend sequence decomposition of positive rates of H7N9 avian influenza virus from April 2013 to April 2019

High-frequency trend sequence decomposition of positive rates of H7N9 avian influenza virus detections from April 2013 to April 2019

Comparisons of predicted values from each model with the true valuesLS-SVM – least square support-vector machines; WPD – wavelet packet decomposition; ARIMA – autoregressive integrated moving average

Spatial autocorrelation analysis of H7N9 avian influenza virus detections in China

Variable	Moran’s I	Z score	P value
From 2013 to 2020	−0.014	0.387	0.350
In 2013	−0.019	0.259	0.398
In 2014	−0.046	−0.362	0.359
In 2015	0.042	1.783	0.037
In 2016	−0.064	−0.992	0.161
In 2017	−0.034	−0.075	0.470
In 2018	−0.007	0.607	0.272
In 2019	−0.021	0.239	0.406
In 2020	—	—	—
In January	−0.008	0.507	0.306
In February	−0.008	0.526	0.299
In March	−0.006	0.563	0.287
In April	−0.004	0.586	0.279
In May	−0.040	−0.210	0.417
In June	−0.045	−0.350	0.363
In July	0.009	1.103	0.135
In August	—	—	—
In September	−0.073	−0.996	0.160
In October	−0.036	−0.164	0.435
In November	−0.046	−0.366	0.357
In December	0.072	2.320	0.010
Before comprehensive animal H7N9 immunisation	−0.020	0.255	0.399
After comprehensive animal H7N9 immunisation	−0.028	0.058	0.477

Optimisation results of low-frequency trend sequence parameters

Decomposition sequence	Hyperparameters σ	Normalisation parameters ɣ
[3,0]	9506	7350.8
[3,1]	2514.5	3946.2
[3,2]	9113.6	9740.8
[3,3]	818.709	7242.1

ARIMA model parameters

Decomposition sequence	ARIMA model parameters
[3,4]	ARIMA (1,0,0)*(2,0,0) [12]
[3,5]	ARIMA (3,0,0)
[3,6]	ARIMA (0,0,0)
[3,7]	ARIMA (3,0,0)

Prediction results of the LS-SVM-ARIMA combined model based on WPD

Date	Real value	[3,0]	[3,1]	[3,2]	[3,3]	[3,4]	[3,5]	[3,7]	Predictive value
05/2019	0	0	0	0.001	0	0	0.001	0	0.002
06/2019	0	0	0	0	0	0	0	0.001	0.001
07/2019	0	0	0.001	0	0	0	0	0	0.001
08/2019	0	0.001	0	0	0.001	0.001	0	0	0.003
09/2019	0	0	0	0	0.001	0	0.001	0	0.002
10/2019	0.001	0.001	0.002	0.001	0.001	0.001	0.002	0.001	0.010
11/2019	0	0.001	0.001	0	0	0.001	0	0	0.003
12/2019	0	0.001	0	0.001	0.001	0	0.001	0.002	0.006
01/2020	0	0	0.001	0	0.001	0.001	0.001	0	0.004
02/2020	0	0.001	0	0	0	0.002	0	0	0.003
03/2020	0	0	0.001	0	0	0	0	0	0.001
04/2020	0	0	0.001	0	0	0.001	0	0	0.002

Evaluation of the prediction results of the different models

Evaluation index	LS-SVM-ARIMA combined model based on WPD	ARIMA model	LS-SVM model	LS-SVM-model ARIMA
MAE	0.024	0.119	0.098	0.077
RMSE	0.020	0.067	0.059	0.041

Situation of H7N9 avian influenza epidemic from April 2013 to April 2020

Time	Province	Number of outbreaks	Poultry	Number of cases	Number of fatalities	Number of animals culled
June 2017	Inner Mongolia	2	Chicken	63,406	37,582	424,197
June 2017	Heilongjiang	1	Chicken	20,150	19,500	16,610
August 2017	Anhui	1	Chicken	1,368	910	74,463
March 2018	Shaanxi	1	Chicken	1,000	810	1,000
April 2018	Shanxi	1	Chicken	812	699	6,374
April 2018	Ningxia	1	Chicken	1,200	585	13,578
May 2018	Liaoning	1	Chicken	11,000	9,000	8,000
May 2018	Ningxia	1	Chicken	3,000	2,210	86,000
March 2019	Liaoning	1	Peacock	9	9	191

Prediction results based on the three models: ARIMA, LS-SVM and ARIMA-LS-SVM

Date	Real value	ARIMA model	LS-SVM model	LS-SVM-ARIMA model
05/2019	0	0.014	0.013	0.009
06/2019	0	0.007	0.006	0.011
07/2019	0	0.004	0.005	0.016
08/2019	0	0.011	0.012	0.012
09/2019	0	0.009	0.008	0.008
10/2019	0.008	0.014	0.013	0.020
11/2019	0	0.017	0.010	0.008
12/2019	0	0.020	0.015	0.016
01/2020	0	0.005	0.006	0.007
02/2020	0	0.010	0.007	0.003
03/2020	0	0.011	0.005	0.005
04/2020	0	0.018	0.012	0.010

eISSN:: 2450-8608
Język:: Angielski

Częstotliwość wydawania:: 4 razy w roku
Dziedziny czasopisma:: Life Sciences, Molecular Biology, Microbiology and Virology, other, Medicine, Veterinary Medicine

Kanał RSS czasopisma

Regional distribution of non-human H7N9 avian influenza virus detections in China and construction of a predictive model

Data publikacji: 05 lip 2021

Zakres stron: 253 - 264

Otrzymano: 04 sty 2021

Przyjęty: 10 cze 2021

DOI: https://doi.org/10.2478/jvetres-2021-0034

Słowa kluczowe
epidemiological survey, aetiological virus detection monitoring, wavelet packet decomposition, LS-SVM-ARIMA model, PSO algorithm

© 2021 Z. Huang et al. published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

Fig. 9

Fig. 10

Spatial autocorrelation analysis of H7N9 avian influenza virus detections in China

Optimisation results of low-frequency trend sequence parameters

ARIMA model parameters

Prediction results of the LS-SVM-ARIMA combined model based on WPD

Evaluation of the prediction results of the different models

Situation of H7N9 avian influenza epidemic from April 2013 to April 2020

Prediction results based on the three models: ARIMA, LS-SVM and ARIMA-LS-SVM

Regional distribution of non-human H7N9 avian influenza virus detections in China and construction of a predictive model

Data publikacji: 05 lip 2021

Zakres stron: 253 - 264

Otrzymano: 04 sty 2021

Przyjęty: 10 cze 2021

DOI: https://doi.org/10.2478/jvetres-2021-0034

Słowa kluczoweepidemiological survey, aetiological virus detection monitoring, wavelet packet decomposition, LS-SVM-ARIMA model, PSO algorithm

© 2021 Z. Huang et al. published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

Fig. 9

Fig. 10

Spatial autocorrelation analysis of H7N9 avian influenza virus detections in China

Optimisation results of low-frequency trend sequence parameters

ARIMA model parameters

Prediction results of the LS-SVM-ARIMA combined model based on WPD

Evaluation of the prediction results of the different models

Situation of H7N9 avian influenza epidemic from April 2013 to April 2020

Prediction results based on the three models: ARIMA, LS-SVM and ARIMA-LS-SVM

Słowa kluczowe
epidemiological survey, aetiological virus detection monitoring, wavelet packet decomposition, LS-SVM-ARIMA model, PSO algorithm