Due to the nature of the imitative techniques used in the SL algorithm, we must producealso an appropriate learning set, as described in Section 3.3, before using it for the decisionmaking. Two additional scenarios, S 4 and S 5 are generated in the same way of the testones (the average daily prices of the training scenarios are showed in Fig. 5.3. Thetwo resulting power flow problems are solved using the FBS-OPF algorithm presentedin Section 4.2 and the outputs are processed to obtain the learning sets as described inSection 3.3.