Judgmental model selection for time series forecasting

We are inviting you to participate in a web-based judgmental forecasting exercise. You are asked to select the best model, based on your judgment, for 32 time series.

The exercise consists of four rounds. Each round will contain 8 series and will be followed by a short questionnaire, while different types of information will be provided on top of the estimated forecasts for the future periods. Detailed instructions will be given prior to each round. You may complete the exercise at once (preferred option) or in multiple instances. Your progress will be saved and you may resume at your convenience.

At the end of the exercise you will be provided with your score. Your final performance will be calculated based on the rank of the models corresponding to your choices, according to the actual (out-of-sample) performance of each model. The participants within the Top-20 will be awarded with Amazon Gift Cards (£50 each).

Research Objectives/Research Questions

  • Evaluate the performance of judgmental model selection process.
  • How does judgmental model selection perform compared to statistical model selection?
  • How we can best design FSSs in order to facilitate the model selection process performed by the forecasters?

Thank you in advance for your participation!

Add a Comment