Bias correction in deep learning systems

This invention describes a method to assess bias in a deep learning system that incorporates quantum components. It involves classifying data with the quantum neural network, then introducing noise into the quantum parts and reclassifying the data. By comparing the initial and noisy classifications, the system determines the quantum components' sensitivity, which can indicate bias.

Why it matters: Filed when quantum computing was highly experimental, the intervening years have seen significant advancements in quantum hardware stability and cloud-based quantum access. This makes the practical implementation and testing of quantum neural networks for bias detection more feasible now.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedQuantum computing hardware, specialized software

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10