Leonardo Banchi
(University of Florence)

The ability to extract general laws from a few known examples depends on the complexity of the problem and on the amount of training data. For learning tasks with "quantum data", e.g. quantum states and channels, the learner's performance is further challenged by the destructive nature of quantum measurements.  Using tools from statistical learning theory we will study which of these tasks are easy in the many-particle setting. For instance, we will show that learning to classify entanglement is challenging, as the problem needs many measurements. On the other hand, learning an observable (like an order parameter) able to identify quantum phases of matter is easy, as, in some cases, only a polynomial number of "simple" measurements is required. Finally, we discuss a possible quantum advantage in learning classical temporal stochastic processes, where quantum models can reach high accuracy with less training data.
 

Zoom registration link:
https://zoom.us/meeting/register/3WwiV13GThmT8R-SMa-eZg