In this chapter, we explored the basic concepts of forecasting techniques. Forecast horizon and several forecast methods were analyzed, all in relation to time series data. A time series constitutes a sequence of observations on a phenomenon. In a time series, we can identify several components—trend, seasonality, cycle, and residual. We learned how to remove seasonality from a time series with a practical example.
Then, the models most widely used to represent time series were addressed—AR, MA, ARMA, and ARIMA. For each model, the basic concepts were analyzed and then a mathematical formulation of the model was provided.
Finally, a Keras LSTM model for time series analysis was proposed. Using a practical example, we saw how we can deal with a time series regression problem with a recurrent neural network model of the LSTM type.
In the next chapter, we will...