Machine learning pipeline for predictions regarding a network

This invention describes a system that predicts future network activity, specifically how much data will move through a network link, device, or virtual network at a given time. It works by automatically trying out a pre-selected group of machine learning models, evaluating each one's performance using standard metrics, and then picking the best model to make the final prediction based on real network data.

Why it matters: Since 2022, the landscape for deploying and managing machine learning systems has evolved significantly. Advances in MLOps tools and cloud-based AutoML services now make the automated training, evaluation, and selection of multiple ML models, as described, much more efficient and practical to implement.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedHeavy software system, ML models, data pipelines

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