Machine Learning Optimization of Quantum Circuit Layouts

The quantum circuit layout (QCL) problem involves mapping out a such that the constraints of device are satisfied. We introduce heuristic, QXX, and its machine learning version, QXX-MLP. latter automatically infers optimal QXX parameter values laid has reduced depth. In order to speed up compilation, before laying circuits out, we use Gaussian function estimate depth compiled circuits. This also informs compiler about region influences most resulting circuit’s present empiric evidence for feasibility method using approximation. QXX-MLP open path feasible large-scale QCL methods.