Hi Donald, suppose we built a SVC model in the online Jupyter Notebook, it can done as follows:
1. Save model to your library directory (refer to line #10-14)
2. Execute the code above, "model_2.m" will then be dumped to your cloud directory /lib.
3. Go to "backtest", load your model from "self.evt.path_lib" (refer to line #15). After that, the model can be used for prediction throughout the backtest process (refer to line #23-25).