Regularization of Riemannian optimization: Application to process tomography and quantum machine learning
The episode discusses a paper on 'Regularization of Riemannian optimization' applied to quantum process tomography and machine learning. The hosts explain how adding regularization terms helps optimize quantum channels by favoring simpler, more parsimonious representations. This leads to faster convergence and better fidelity in tomography, and suggests using this method to find the minimum complexity needed for quantum classification tasks.