Summary
The focus of this chapter has been heavily on ARIMA modeling. This is because ARIMA models, by their very definition, focus on the simple core mathematical concepts that effectively define what a time series is. Focusing on ARIMA modeling allowed us to easily introduce and understand those math concepts. Specifically, we have had to learn about the following:
- How the time element of time series data provides a natural ordering to the data and introduces auto-correlation between the values from nearby time points
- How ARIMA models are built up from simple mechanisms that capture auto-correlation patterns in data
- How ARIMA models can be expressed in simple math terms using the lag operator
- How ARIMA models can be extended to include exogenous factors and seasonality
- How ARIMA models are built in practice
- How machine learning algorithms can also be applied to time series data; in particular, how deep learning auto-regressive neural networks incorporate...
