Lab 4: Introduction to Training Quantum Circuits

• Download the notebook: [en] [ja]

In this lab, you will learn how to train circuit-based variational models, using different training techniques and see restrictions the models have and how they might be overcome.

Computing Expectation Values:

• Graded Exercise 4-1: By matrix multiplication

• Graded Exercise 4-2: By simulation

Training A New Loss Function:

• Graded Exercise 4-3: Define the Hamiltonian

• Graded Exercise 4-4: Use the SPSA optimizer to find the minimum

Natural Gradients:

• Exploratory Exercise: Natural Gradients and Barren Plateaus

Suggested resources

• Read Maria Schuld, Ville Bergholm, Christian Gogolin, Josh Izaac, Nathan Killoran on Evaluating analytic gradients on quantum hardware

• Read James Stokes, Josh Izaac, Nathan Killoran, Giuseppe Carleo on Quantum Natural Gradient

• Read Julien Gacon, Christa Zoufal, Giuseppe Carleo, Stefan Woerner on Simultaneous Perturbation Stochastic Approximation of the Quantum Fisher Information

• Read Jarrod R. McClean, Sergio Boixo, Vadim N. Smelyanskiy, Ryan Babbush, Hartmut Neven on Barren plateaus in quantum neural network training landscapes

• Read M. Cerezo, Akira Sone, Tyler Volkoff, Lukasz Cincio, Patrick J. Coles on Cost Function Dependent Barren Plateaus in Shallow Parametrized Quantum Circuits