TIME SERIES MODEL COMPARISON FOR SALEM TO PREDICT TRENDS IN TYPHOID CASE

Author (s) : S.NITHYA and G.KARTHIKA

---

Abstract :
Time series models have been used with varying degrees of success to predict diseases trends. In this study, typhoid incidences in Salem were compared, appraised, and examined using time series models of cases, along with created trends. Using three time series models, autoregressive integrated moving average (ARIMA), support vector machine regression, and exponential smoothing, the data gathered from the Government of India's integrated diseases surveillance programs from 2017 to 2018 was subjected to time series analysis. Low numbers indicate a superior model. The performance in predicting the number of instances is evaluated as root mean square error and mean absolute error. With values of root mean square error and mean absolute error for exponential smoothing, support vector regression, and ARIMA of 13.30 and 4.51, 16.49 and 10.91, and 17.39 and 11.42, respectively, the exponential smoothing model beat the other two models. The findings show that in order to reverse the current trend, rigorous sanitation-related efforts are required. More real-time data sets and the ensemble method will be used in future investigations.

Keywords : Autoregressive integrated moving average model, exponential smoothing model, support vector machine regression model, time series model, typhoid.

Conference Date : 30/08/2019

Pages : 128

Paper ID : 19126