Machine Learning and Quantum

In this talk, I will first talk about “ machine learning of quantum problems”. Since quantum problems are governed by well-defined physical rules, they are good platforms to investigate explainable machine learning. Here I will give one example of learning potential-to-density mapping by the recurrent neural network, from which we can extract the Schrodinger equation. Secondly, I will talk about “ machine learning for quantum experiments”. Especially, I will describe how to use the idea of active learning to optimize quantum control, which requires minimal number of experimental data. Finally, I will talk about “ machine learning on quantum compute”. I will bring together the concept information scrambling and quantum machine learning. I will show that the tripartite information developed for quantifying information scrambling can be used to diagnose the training dynamics of a quantum neural network.