Skip to content

Overview

The following sections provide an overview of the structure of forecasting tasks TIM executes, the inputs that are expected and the outputs that will be calculated.

TIM Forecasting structure

The image below shows the TIM Forecasting process.

image.png

In order to create a forecast, TIM requires the following inputs:

  • Data: historical data used to build and train the model(s);

  • Configuration: mathematical settings and other relevant parameters. Most configuration settings are optional, with exception of the prediction horizon; and

  • Model Zoo: previously built models that can be used in the model building and/or forecasting process. This is an optional parameter.

These inputs are processed by the TIM engine during the model building phase. The following is returned from this phase:

  • Model Zoo: a new Model Zoo that was merged from the input Model Zoo (if any) and the newly built models;

  • Forecasts: a tabular output encapsulating all backtesting and production forecasts, as well as insights into the model(s); and

  • Messages: errors, warnings and other informative messages about your data and/or the forecasting process.