Machine Learning for Time-Series with Python

5

Forecasting with Moving Averages and Autoregressive Models

This chapter is about time-series modeling based on moving averages and autoregression. This comprises a large set of models that are very popular in different disciplines, including econometrics and statistics. We'll discuss autoregression and moving averages models, along with others that combine these two, such as ARMA, ARIMA, VAR, GARCH, and others.

These models are still held in high esteem and find their applications. However, many new models have since sprung up that have been shown to be competitive or even outperform these simpler ones. Within their main application, however, in univariate forecasting, simple models often provide accurate or accurate enough predictions, so ...