This is a LightGBM time-series forecasting model. LightGBM is a gradient boosting decision tree framework developed by Microsoft. It works by recursively partitioning the feature-space into hyperrectangles and utilising the mean (or median) of the target in the specific hyperrectangle as prediction. Every one step recursion is made to reduce the prediction errors of the previous model iteration. One of the advantages with LightGBM over other gradient boosting decision tree frameworks is its efficiency and the ability to predict quantile distributions. 

The asset provides a user interface where you can upload a train set and a set to predict on. The prediction is then displayed in a chart and can be downloaded from the user-interface. It also exposes the rpc Predict() to be able to be called from another service. Here is a video demonstration. Please refer to this readme for more information about how to use and install.

This project has received funding from the European Union's Horizon 2020 research and innovation programme within the framework of the I-NERGY Project, funded under grant agreement No 101016508

Website: https://www.rebase.energy/