Introduction to Time Series Modeling; Providing Applied Patterns

Series modeling involves four general phases when relying on data analysis: diagnosis, analysis, composition, and validation. Unfortunate judgment in each of the four phases will lead to a less optimal model. In the analysis of a time series, there are five general components, that all of which may be present in a time series or vice versa. Three components can be described as systematic: long-term, periodic, and cyclical trends. Episodic events and random changes are components that reflect sources of non-systematic change. In each case, a practical example is provided. For example, the use of this modeling to predict the annual peak of discharges for the northwestern branch of the Anakostia river in Hitcheville, Maryland for water years 1939 to 1988, showed that the logistics model may be a reasonable model to show the middle section of the trend of long-term.