I have built an ARIMAX model in python for predicting a time series. After presenting my findings ive been asked what robustness tests i have used. My skillset is more on the python side. I only have basic statistics training and i dont know how to answer this question. I can obviously review the model output (coefficients, p-values, errors) but my understanding is that robustness refers to:

statistics that yield good performance when data is drawn from a wide range of probability distributions that are largely unaffected by outliers or small departures from model assumptions in a given dataset. (https://www.thoughtco.com/what-is-robustness-in-statistics-3126323)

How do i do this?

  1. If i run the arimax using various different train/test subsets of my time series, would that qualify as a robustness test? How many times should i do this?
  2. Are there any standardised robustness tests for an arimax prediction? What are they?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.