Reducing Measurement Cost with Informationally Complete Measurements

Reducing Measurement Cost with Informationally Complete Measurements

Seminar author:Dani Cavalcanti

Event date and time:04/09/2026 02:30:pm

Event location:GIQ Seminar Room

Event contact:

Estimating expectation values of observables is a central challenge in quantum computing, particularly for quantum chemistry applications. Standard approaches based on Pauli decomposition or Classical Shadows face limitations at scale, including exponential growth in measurement cost and difficulties handling correlated or noisy measurements.

This talk presents a framework for observable estimation using informationally complete measurements and optimised dual frames, of which Classical Shadows is a special case. We introduce two methods — Locally Optimal Dual Frames and Dual Optimisation with Tensors  — that significantly reduce estimator variance through classical post-processing alone, requiring no changes to the quantum hardware measurements. Both methods are benchmarked on molecular Hamiltonians demonstrating substantial improvements over Classical Shadows.