I am happy to present our new work 🎆:
https://t.co/jzUEFjqKi1
Thanks to all my collaborators @MatiasBilkis, Mauricio Matera, Lorena Rebón, @fervilarino for all the effort in this work and @CVC_UAB for welcoming me with the internship Elena Maseras.💎
7- As expected, this example shows big improvements over standard model-free calibration methods such as Q-learning 🤯. Moreover, this scheme presents the possibility of generalizing more the individual stages, to further improve the calibration of complex systems 🎛️.
I am happy to present our new work 🎆:
https://t.co/jzUEFjqKi1
Thanks to all my collaborators @MatiasBilkis, Mauricio Matera, Lorena Rebón, @fervilarino for all the effort in this work and @CVC_UAB for welcoming me with the internship Elena Maseras.💎
6- After the change is detected, we use a few experiments to fix a simple model, that helps the Reinforcement Learning agent find the optimal configuration.