data that has been passed into the model is stationary or not ?
Yes because stationary data is supposedly more independent which means a more random error which follows a normal distribution. That is why when doing time series analysis you might want to take derivatives of the data until you get a stationary sample
Обсуждают сегодня