In this section, we will conceptually explain the following special characteristics of time series data that requires its special mathematical treatment:
- General trend
- Seasonality
- Cyclical movements
- Unexpected variations
Most time series has of one or more of the aforementioned internal structures. Based on this notion, a time series can be expressed as xt = ft + st + ct + et, which is a sum of the trend, seasonal, cyclical, and irregular components in that order. Here, t is the time index at which observations about the series have been taken at t = 1,2,3 ...N successive and equally spaced points in time.
The objective of time series analysis is to decompose a time series into its constituent characteristics ...