Training neural networks using evolution based strategies and novelty search
US11068787B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Dec 14, 2018 |
| Grant date | Jul 20, 2021 |
| Priority date | — |
| Expiry date | May 4, 2039 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG06N5/043
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
Systems and methods are disclosed herein for selecting a parameter vector from a set of parameter vectors for a neural network and generating a plurality of copies of the parameter vector. The systems and methods generate a plurality of modified parameter vectors by perturbing each copy of the parameter vector with a different perturbation seed, and determine, for each respective modified parameter vector, a respective measure of novelty. The systems and methods determine an optimal new parameter vector based on each respective measure of novelty for each respective one of the plurality of modified parameter vectors, and determine behavior characteristics of the new parameter vector. The systems and methods store the behavior characteristics of the new parameter vector in an archive.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.