Automated compilation of probabilistic task description into executable neural network specification

This invention describes a system that automatically builds a neural network from a high-level statistical description of a problem. You provide a model explaining how data is generated, and the system converts this into a blueprint, then into a neural network designed to solve a specific prediction task. The claims narrow this to creating a "discriminative message passing network," which is a type of neural network focused on making predictions.

Why it matters: Filed in 2016, this invention predates the widespread maturity of neural architecture search (NAS) and advanced probabilistic programming frameworks. Today, the tools and techniques for automatically generating and optimizing neural network structures from high-level descriptions are far more developed, making the automated compilation of probabilistic models into efficient neural networks a much more feasible and impactful endeavor for tasks like complex inference.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedMachine learning software, probabilistic programming expertise.

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