Data and code for the paper "Data-Driven Conditional Robust Optimization" 

Structure:
- code: there are three sub-folders:
-- generator: a python file used to generate random data
-- solver: optimization code, takes trained parameters as input
-- train_nn: code to train nn in combination with the conditional Deep SVDD code 

- scripts: shell scripts used to run experiments, which in particular give parameter choices
--log: logs from all the runs get saved here

-path: the models generated from the training get saved here
-- data: the data used for portfolio optimization is generated and saved here when the shell scripts are run.