Indexing

Comparing SARIMA and Dynamic Seasonal Model: Application to Acute Encephalitis Syndrome (AES) for Gorakhpur, India

Author: Pooja Kushwaha and Richa Vatsa

PDF Download PDF

Abstract

In this paper, we applied a Bayesian dynamical seasonal modelling of count data. Their usefulness is illustrated by their application to Acute Encephalitis Syndrome (AES) cases from Gorakhpur regions and by comparing them with the widely used seasonal autoregressive integrated moving average (SARIMA) models for seasonal modelling. The outbreak of encephalitis causes many deaths and long-term disabilities among children and young adults. We considered the AES case data of Gorakhpur from (Jan-12 to Nov-17). We focus on the case of response variables following a Poisson distribution, concentrating on the dynamical seasonal harmonic model. The study helps the policy maker, future disease spread and a better understanding of high-risk months, which may be associated with AES cases. Prior knowledge of the disease outbreak is a main and essential step for policymakers to minimise the disease risk and mortality of children, and enhance health services, vaccination programmes, and other public health initiatives.

Keywords

AES, Bayesian estimation, dynamic seasonal harmonic model, cross-validation, SARIMA, WAIC

Conclusion

In our analysis, we found a clear seasonal peak in September. August and October months are also positively associated with the disease. The dynamic seasonal harmonic Poisson model modelled non-stationarity in seasonal data better than SARIMA models. However, the limit of our study is that it is based only on trends and seasonal effects, with no other explanatory variables included in the study. But this may be seen as an opportunity. In all circumstances, one wants to know the disease pattern and its association with different months. The dynamic seasonal model may provide a better understanding than the SARIMA models.

References

-

How to cite this article

Pooja Kushwaha and Richa Vatsa (2023). Comparing SARIMA and Dynamic Seasonal Model: Application to Acute Encephalitis Syndrome (AES) for Gorakhpur, India. Biological Forum – An International Journal, 15(3): 534-542.