This a repository from my learnings on time series forecasting in Python.
The notebooks in this repository go over:
- Introduction to Facebook's Prophet forecasting package
- Variables and data types intro (skip if you know this)
- Prophet forecasting evaluation metrics (MAPE, MAE, RMSE)
- Descriptive statistics with
statsmodels
- Working with dates and times
- Looking at time resampling and how this can be used to split data
- Time shifting - random walks, and other methods
- Rolling and expanding time window methods
- Visualising time series
- Changing seasonality
- Introduction to
statsmodels
- ETS (Error Trend Seasonality) modelling
- Exponential Weighted Moving Averages (EWMA)
- Holts-Winters Exponential Smoothing
- Introduction to forecasting
- ACF and PACF plotting
- Autoregression with stats models
- Descriptive time series and testing methods
- Choosing ARMA orders
- ARMA, ARIMA and SARIMA Autoregression models
- Season Arima with exogenous variables
- Vector Autoregressions
- Prophet Python forecasting