Figure 2
From: Pareto optimization with small data by learning across common objective spaces
The PE workflow. The inverse machine learning model can be queried with arbitrary preference vectors \(W_q\) by the DM to get predicted solutions \(X_q\) in the PS.
