Contents
Why is stationarity important for forecasting?
Stationarity is an important concept in the field of time series analysis with tremendous influence on how the data is perceived and predicted. The best indication of this is when the dataset of past instances is stationary. For data to be stationary, the statistical properties of a system do not change over time.
Which Neural Networks is ideal for time series data?
Convolutional Neural Networks or CNNs are a type of neural network that was designed to efficiently handle image data. The ability of CNNs to learn and automatically extract features from raw input data can be applied to time series forecasting problems.
Are there predictable patterns in a stationary time series?
In general, a stationary time series will have no predictable patterns in the long-term. Time plots will show the series to be roughly horizontal (although some cyclic behaviour is possible), with constant variance. Figure 8.1: Which of these series are stationary?
When to use seasonal differencing or stationarity?
However, if the data have a strong seasonal pattern, we recommend that seasonal differencing be done first, because the resulting series will sometimes be stationary and there will be no need for a further first difference. If first differencing is done first, there will still be seasonality present.
How to determine if differencing is required for stationarity?
One way to determine more objectively whether differencing is required is to use a unit root test. These are statistical hypothesis tests of stationarity that are designed for determining whether differencing is required.
How does stationarity and differencing affect A10 sales?
The transformation and differencing have made the series look relatively stationary. Figure 8.3: Logs and seasonal differences of the A10 (antidiabetic) sales data. The logarithms stabilise the variance, while the seasonal differences remove the seasonality and trend.