Treatment on outliers in UBJ-SARIMA models for forecasting dengue cases on age groups not eligible for vaccination in Baguio City, Philippines

Dengue vaccination has become a breakthrough in the fight against dengue infection. This is however not applicable to all ages. Individuals from 0 to 8 years old and adults older than 45 years old remain susceptible to the vector-borne disease dengue. Forecasting future dengue cases accurately from...

Full description

Saved in:
Bibliographic Details
Main Authors: Magsakay, Clarenz B., De Vera, Nora U., Libatique, Criselda P., Addawe, Rizavel C., Addawe, Joel M.
Format: Conference Proceeding
Language:English
Subjects:
Adults
Age
Age groups
Dengue fever
Forecasting
Immunization
Mathematical models
Outliers (statistics)
Transformations
Viral diseases
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Dengue vaccination has become a breakthrough in the fight against dengue infection. This is however not applicable to all ages. Individuals from 0 to 8 years old and adults older than 45 years old remain susceptible to the vector-borne disease dengue. Forecasting future dengue cases accurately from susceptible age groups would aid in the efforts to prevent further increase in dengue infections. For the age groups of individuals not eligible for vaccination, the presence of outliers was observed and was treated using winsorization, square root, and logarithmic transformations to create a SARIMA model. The best model for the age group 0 to 8 years old was found to be ARIMA(13,1,0)(1,0,0)12 with 10 fixed variables using square root transformation with a 95% winsorization, and the best model for the age group older than 45 years old is ARIMA(7,1,0)(1,0,0)12 with 5 fixed variables using logarithmic transformation with 90% winsorization. These models are then used to forecast the monthly dengue cases for Baguio City for the age groups considered.
ISSN:0094-243X
1551-7616
DOI:10.1063/1.5012247