decomposition_of_time_series_data.md

##Time Series Decomposition##

A time series, TS, is usually broken down into 4 components:

1. Trend component

   • Reflects the long-term progression/behaviour of a TS (secular variation)
   • A trend is present in a TS if there is a persistent increase or decrease occuring in a certain direction
   • Does not need to be linear

2. Cyclical component

   • Reflects repeating, non-periodic fluctuations present within a TS
   • "The duration of these fluctuations is usually of at least two years." - Wikipedia, Decomposition of time series

3. Seasonal component

   • Reflects the seasonality (seasonal variation) present in a TS
   • A seasonal pattern exists when a time series is influenced by seasonal factors
   • Seasonality occurs over a fixed and known period of time (e.g. the 1/4 of a year, the month or the day of the weeek)

4. Residual (aka "noise"?) component

   • Represents the remains of a TS after all its other components have been removed from it