The Deep Signal Library has added advanced methods for controlling exactly what data sets you would like to use for training your machine learning model. This takes the previous version 1.1 additions a step further and gives absolute control to the user for what data sets they want to use for training their machine learning model. There are four new methods that allow the trader to add data sets that have reached a long or short profit target as well as data sets that failed to reach a long or short profit target. The machine learning library uses supervised learning and needs both types of data to accurately train your model.
Please see our documentation on how to use this new functionality at the Creating Custom Data Sets online help guide.