A disease outbreak detection system using autoregressive moving average in time series analysis

Disease outbreak detection is an immensely beneficial yet a very delicate procedure. Its output can potentially provide immeasurable help and save countless lives in preventing further problems. Thus, the method must be as accurate and precise as possible. This paper discusses the Disease Outbreak Detection System, which is an online system developed to aid public health workers as they resolve this problem. Specifically, it helps epidemiologists analyze the behavior of a certain disease outbreak by providing them prediction values for a specific time interval. The system is able to perform such feature with the aid of R software which performs computations of Time Series analysis using Autoregressive Moving Averages (ARMA) Model to generate values based on the present condition of the outbreak. These generated values will serve as basis to know how the outbreak will turn out, thus giving the epidemiologist sufficient time to respond to major public health threats and formulate preventive measures to control and solve the outbreak. The tool was developed for the Philippine setting, specifically for the use of outbreak monitoring agencies such as the National Epidemiological Center (NEC), and thus uses Philippine health data which basically comes from two major sources: surveys and censuses, as well as from administrative records of health and health related agencies.