Time Series: Decomposition Theory (TS E3)

Time series can be broken down into components for modeling and understanding purposes. These are typically trend, cycle, and season as well as the error term. The only difference between cycle and season is the number of cycles that occur over a time period. Cycles are usually annually or longer whereas seasons are usually less than a year. You can also have multiple seasons in one model. This video is going to cover the basic theory behind this decomposition and how they work together to model a time series. Future videos will give examples on how this would work with real data. TS E1: Business vs Statistical Analytics: Concept Overview https://youtu.be/hvxQphdRzUQ TS E2: Time Series Intro: Stochastic Processes and Structure https://youtu.be/fXcyLngs6xw Support this channel: https://streamlabs.com/dimitribianco