Robust two-stage influenza prediction model considering regular and irregular trends

Murayama, Taichi; Sato, Nobuyuki; Fujita, Sumio; Wakamiya, Shoko; Aramaki, Eiji

doi:10.1371/journal.pone.0233126

Cited by 5 publications

(6 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our results showed the correlations for the LCI and LCIPP demonstrate external validity, as they were both associated with constructs that could be explained by previous research. This study joins a growing body of literature that uses web-based search queries to track public health (eg, Murayama et al [32]), and the LCIPP adds the dimension of using prefecture-level infection rates to control for expected outcomes. Therefore, this study was able to effectively quantify public concern at a deeper level, which we propose is explained by the collectivistic psychological tendencies of a society.…”

Section: Effectiveness Of Search Queries As Public Concern Indicatorsmentioning

confidence: 99%

Measuring Public Concern About COVID-19 in Japanese Internet Users Through Search Queries: Infodemiological Study

Gao¹,

Fujita²,

Sato³

et al. 2021

JMIR Public Health Surveill

Self Cite

View full text Add to dashboard Cite

Background COVID-19 has disrupted lives and livelihoods and caused widespread panic worldwide. Emerging reports suggest that people living in rural areas in some countries are more susceptible to COVID-19. However, there is a lack of quantitative evidence that can shed light on whether residents of rural areas are more concerned about COVID-19 than residents of urban areas. Objective This infodemiology study investigated attitudes toward COVID-19 in different Japanese prefectures by aggregating and analyzing Yahoo! JAPAN search queries. Methods We measured COVID-19 concerns in each Japanese prefecture by aggregating search counts of COVID-19–related queries of Yahoo! JAPAN users and data related to COVID-19 cases. We then defined two indices—the localized concern index (LCI) and localized concern index by patient percentage (LCIPP)—to quantitatively represent the degree of concern. To investigate the impact of emergency declarations on people's concerns, we divided our study period into three phases according to the timing of the state of emergency in Japan: before, during, and after. In addition, we evaluated the relationship between the LCI and LCIPP in different prefectures by correlating them with prefecture-level indicators of urbanization. Results Our results demonstrated that the concerns about COVID-19 in the prefectures changed in accordance with the declaration of the state of emergency. The correlation analyses also indicated that the differentiated types of public concern measured by the LCI and LCIPP reflect the prefectures’ level of urbanization to a certain extent (ie, the LCI appears to be more suitable for quantifying COVID-19 concern in urban areas, while the LCIPP seems to be more appropriate for rural areas). Conclusions We quantitatively defined Japanese Yahoo users’ concerns about COVID-19 by using the search counts of COVID-19–related search queries. Our results also showed that the LCI and LCIPP have external validity.

show abstract

Section: Effectiveness Of Search Queries As Public Concern Indicatorsmentioning

confidence: 99%

Measuring Public Concern About COVID-19 in Japanese Internet Users Through Search Queries: Infodemiological Study

Gao¹,

Fujita²,

Sato³

et al. 2021

JMIR Public Health Surveill

Self Cite

View full text Add to dashboard Cite

show abstract

“…Our model is motivated by the idea that search query data are useful features for forecasting non-seasonal parts of flu data. This concept originates from a previous study [35], which reported that the flu forecasting accuracy is improved by splitting the forecasting part from the historical ILI data and search query data. The model architecture is presented in Fig.…”

Section: Model Structurementioning

confidence: 99%

“…-Two-stage: The Two-stage model [35], composed of long short-term model and AR model, was developed inspired by a similar idea to ours, in that the usefulness of the input data differs; historical ILI data and search query data are useful for forecasting the seasonality and trend, respectively. For the multi-step-ahead forecast, we extended the two-stage model to the encoderdecoder architecture.…”

Section: Comparative Models -Grumentioning

confidence: 99%

“…Most of these approaches do not consider the characteristics of the data types, but simply simultaneous inputting, when learning the model. Several previous models [37,35] based on statistical methods have been developed to exploit the characteristics of the different input data. However, these studies exhibit certain disadvantages, such as the necessity of long training terms or a small degree of improved accuracy.…”

Section: B Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Single Model for Influenza Forecasting of Multiple Countries by Multi-task Learning

Murayama¹,

Wakamiya²,

Aramaki³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

The accurate forecasting of infectious epidemic diseases such as influenza is a crucial task undertaken by medical institutions. Although numerous flu forecasting methods and models based mainly on historical flu activity data and online user-generated contents have been proposed in previous studies, no flu forecasting model targeting multiple countries using two types of data exists at present. Our paper leverages multi-task learning to tackle the challenge of building one flu forecasting model targeting multiple countries; each country as each task. Also, to develop the flu prediction model with higher performance, we solved two issues; finding suitable search queries, which are part of the usergenerated contents, and how to leverage search queries efficiently in the model creation. For the first issue, we propose the transfer approaches from English to other languages. For the second issue, we propose a novel flu forecasting model that takes advantage of search queries using an attention mechanism and extend the model to a multi-task model for multiple countries' flu forecasts. Experiments on forecasting flu epidemics in five countries demonstrate that our model significantly improved the performance by leveraging the search queries and multi-task learning compared to the baselines.

show abstract

“…Seasonal influenza, caused by the influenza virus, is an acute respiratory disease characterized by the sudden onset of fever, headache, cough, rhinitis, and muscle and joint pain [1,2]. Influenza can be classified into four types: influenza A virus, influenza B virus, influenza C virus, and influenza D virus, among which influenza virus types A and B circulate and cause seasonal influenza [3].…”

Section: Introductionmentioning

confidence: 99%

A new hybrid model SARIMA-ETS-SVR for seasonal influenza incidence prediction in mainland China

Zhao,

Zhang

2023

J Infect Dev Ctries

View full text Add to dashboard Cite

Introduction: Seasonal influenza is a serious public health issue in China. This study aimed to develop a new hybrid model for seasonal influenza incidence prediction and provide reference information for early warning management before outbreaks. Methodology: Data on the monthly incidence of seasonal influenza between 2004 and 2018 were obtained from the China Public Health Science Data Center website. A single seasonal autoregressive integrated moving average (SARIMA) model and a single error trend and seasonality (ETS) model were built. On this basis, we constructed SARIMA, ETS, and support vector regression (SARIMA-ETS-SVR) hybrid model. The prediction performance was determined by comparing mean absolute error (MAE), mean square error (MSE), mean absolute percentage error (MAPE), and root mean square error (RMSE) indices. Results: The optimum SARIMA model was SARIMA (0,1,0) (0,0,1)12. Error trend and seasonality (ETS) (M,A,M) was the SARIMA optimal model. For the fitting performance, the SARIMA-ETS-SVR hybrid model achieved the lowest values of MAE, MSE, and RMSE, in addition to the MAPE. In terms of predictive performance, the SARIMA-ETS-SVR hybrid model had the lowest MAE, MSE, MAPE, and RMSE values among the three models. Conclusions: The study demonstrated that the SARIMA-ETS-SVR hybrid model provides better generalization ability than a single SARIMA model and a single ETS model, and the predictions will provide a useful tool for preventing this infectious disease.

show abstract

Robust two-stage influenza prediction model considering regular and irregular trends

Cited by 5 publications

References 26 publications

Measuring Public Concern About COVID-19 in Japanese Internet Users Through Search Queries: Infodemiological Study

Measuring Public Concern About COVID-19 in Japanese Internet Users Through Search Queries: Infodemiological Study

Single Model for Influenza Forecasting of Multiple Countries by Multi-task Learning

A new hybrid model SARIMA-ETS-SVR for seasonal influenza incidence prediction in mainland China

Contact Info

Product

Resources

About