Our approach

Any data science analysis involves not only applying some specific algorithm of interest, but also a data preparation step and data quality checks. Our approach is based on including these steps into the forecast evaluation process allowing it to comply with more general methodologies such as the Microsoft TDSP, CRISP-DM, SEMMA, etc. The forecast evaluation framework proposed below summarizes our experience of implementing forecasting projects in various settings.

The major steps of the framework are outlined in the forecast evaluation workflow. Subsequent sections provide more details on on the tools chosen, the rationale behind them, and examples in R.

To cite this website, please use the following reference:

Sai, C., Davydenko, A., & Shcherbakov, M. (date). The Forvision Project. Retrieved from https://forvis.github.io/

_{© 2020 Sai, C., Davydenko, A., & Shcherbakov, M. All Rights Reserved. All rights reserved. Short sections of text, not exceed two paragraphs, may be quoted without explicit permission, provided that full acknowledgement is given.}