qml.kernels.mitigate_depolarizing_noise

mitigate_depolarizing_noise(K, num_wires, method, use_entries=None)

Estimate depolarizing noise rate(s) using on the diagonal entries of a kernel matrix and mitigate the noise, assuming a global depolarizing noise model.

Parameters:

• K (array[float]) – Noisy kernel matrix.

• num_wires (int) – Number of wires/qubits of the quantum embedding kernel.

• method ('single' | 'average' | 'split_channel') –

  Strategy for mitigation

  • 'single': An alias for 'average' with len(use_entries)=1.

  • 'average': Estimate a global noise rate based on the average of the diagonal entries in use_entries, which need to be measured on the quantum computer.

  • 'split_channel': Estimate individual noise rates per embedding, requiring all diagonal entries to be measured on the quantum computer.

• use_entries (array[int]) – Diagonal entries to use if method in ['single', 'average']. If None, defaults to [0] ('single') or range(len(K)) ('average').

Returns:

Mitigated kernel matrix.

Return type:

array[float]

Reference:

This method is introduced in Section V in arXiv:2105.02276.

Example:

For an example usage of mitigate_depolarizing_noise please refer to the PennyLane demo on the kernel module or the postprocessing demo for arXiv:2105.02276.

code/api/pennylane.kernels.mitigate_depolarizing_noise

 

Download Python script

 

Download Notebook

 

View on GitHub