In stationary, if we want to make forecasting, we have to make our data stationary(On classical methods), I get that, but If every data point is independent of each other, how can we make the prediction of what will be the value of the next time step? I mean there is no pattern, no association, nothing. How a model can estimate what will be the next value?

And while this is the case, Imagine I have data that every time stamp has different values (their meaning is completely independent of each other), and this is a classification problem. (I want to classify each row), based on the assumption of "stationary time series", my values are already stationary, so, can I easily convert this data to the "time series data" and forecast it?


  • 1
    $\begingroup$ Intuitively, If your values are stationary, Then it would be following "reversion to mean", So if the last value is very high/very low when compared to mean, Then the next value will be much closer to mean. Isn't that useful for a model to estimate its next value? $\endgroup$ Oct 14, 2022 at 12:40
  • $\begingroup$ Thank you so much for your response, but i am sorry that i couldn't understand what you mean by "Then it would be following "reversion to mean" Is it possible for you to illustrate or give some examples, please? $\endgroup$
    – Canovich
    Oct 14, 2022 at 12:57


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.