Time Series Decomposition
Overview
The classical time series decomposition model, models a time series as a sum of a trend, seasonal factors, and a
random noise term.
{% y_t = trend_t + seasonality_t + trendReversion_t + \epsilon_t %}
Technically speaking, the seasonality can be considered as part of the trend. However, it is often
separated out so that it can be analyzed separately. Additionally, whereas trend is usually considered to
be the component of the time series which makes it non-stastionary, seasonality does not affect the
stationarity
of the time series.
Components
