Closed AlexMV12 closed 3 years ago
Sorry, this list is only for open source tools. I would like to exclude commercial solutions.
Hello Max,
I'm sorry, why do you think TIMEX is a commercial solution? The code is open source. Is it something related to the license? Thank you
Sorry, the "-as-a-service" part made me think that its commercial solution.
I see, no problem. Indeed, in recent years a lot of "as-a-service" platforms for different tasks, i.e. text translation has been deployed. However, also open-source/self-hostable platforms has been created and released. As of today, there are few time-series forecasting-as-a-service platforms. Probably the biggest one is Amazon AWS Forecast which, however, is absolutely proprietary.
TIMEX is an attempt to release a platform which can be easily adapted to work similarly to AWS Forecast, but, obviously, with the code open-sourced and released. It is different with respect to other libraries in the sense that it is aimed at creating as-a-service websites/platforms, instead of providing helper functions to ease some tasks of time-series processing (for which excellent libraries already exist).
Hello,
First of all, thank you for your time and for this project. It helped us very much in individuating works similar to the one we intended to do.
TIMEX is a Python framework for time-series-forecasting-as-a-service. Its main goal is to provide a way to offer users the possibility to access a fully-automated pipeline for time-series ingestion, pre-processing, and forecasting. Optionally, a visualization part, built on Dash is offered.
One example of the capabilities of TIMEX is this website, which we made available to the public for Covid-19 forecasting in Italy.
This site could be modified to allow users to specify/upload the input data, which models to employ, using a simple JSON file, transforming it in a customizable platform for time-series-forecasting-as-a-service.
We would like to have feedback from the community, and we would be grateful if you could include TIMEX in your repository. In the future a paper on TIMEX will be made public.
Thank you very much,
Alessandro and Manuel Politecnico di Milano