Random Forest algorithm to forecast the hourly power generation of PV plant over time, using time series data and multiple features. Note that the data is deducted due to confidentiality but a version of the data file is published for users to understand the data format. You can see some example predictions from the testing stage.
The error metrics were as follows:
-r2 = 0.9389
-MSE: 0.4072
-MAE: 0.316
-RMSE: 0.638