Noise-boosted back propagation and deep learning neural networks

This invention describes a computer system that uses deep learning, a type of artificial intelligence, to learn and adapt. It works by processing data through layers of virtual units, comparing the results to what it expects, and then adjusting itself based on the differences. Uniquely, this system intentionally adds random or chaotic numerical disturbances, like noise, into its learning process without any conditions, specifically to make it learn faster.

Why it matters: Filed before the explosion of extremely large and complex deep learning models, the problem of training speed and efficiency has become exponentially more critical by 2026. The availability of advanced AI hardware and mature software frameworks makes implementing and testing such a system, which claims to speed up learning, significantly more feasible and impactful now.

Status: ExpiredProtection ended and was never renewed
How hard to build: SpecializedSoftware logic, specialized AI hardware, complex algorithms

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