Time Series Forecasting in Python

Disclaimer

The contents here are the notes taken by me while I was reading the book Time Series Forecasting in Python by Marco Peixeiro. The book can be found online at this link. The book is also available at Oreilly online library if your organization has a subscription to Oreilly.

The write up here is merely for notes so that I can revisit later on demand. The rights to the contents of the books remain with the author.

Back To Chapter List

My Notes from Chapter-01: Basics

1. First, we will learn to make simple forecasts. Then, we will use moving average model and the autoregressive model to make forecasts.
   1. These will be foundation for complex modeling technique that will allow us to account for diferent components of real world time series, like, non stationarity, seasonality and impact of exogenous variables.
2. Later on, we will move on to Deep Learning methods
3. What is time series?: A set of data points ordered in time.
   1. Time series data are usually equally spaced.
4. Components of Time Series:
   • Trend Component: The slow moving change of a time series that defines the characteristic. Shows the general direction of the time series.
   • Seasonal Component: Repititive component that cycles over certain period.
   • Residual Component:
5. Forecasting:
   1. If you observe seasonality, SARIMA model would be better as it accounts seasonals effects.
6. Time Series forecasting different from other regression tasks:
   • Time series have an order: In other tasks, it does not matter when x leads to y. It should hold for any time that x occurs. That is not the case in timeseries.
   • TIme series sometimes have no features, just time and the value of concerned metric/variable.