Imitation learning for machine learning systems with synthetic data generators
AssigneeD5Ai LLC
InventorJames K. Baker
Life of this patent
Estimating the term…
Abstract
Computer systems and methods cooperatively train multiple generators and a classifier. Cooperative training includes: training, through machine learning, the multiple generators such that each generator is trained according to a first objective to output examples of a designated classification category; training, through machine learning, the classifier to determine, for each generated by the multiple generators, which of the multiple generators generated the example; and back-propagating partial derivatives of an error cost function from the classifier to the multiple generators.
Claims45
A machine-learning method performed by a computer system having one or more programmed processor cores, the method comprising cooperatively training multiple generators and a classifier, wherein cooperatively training the multiple generators and the classifier comprises:
- training, through machine learning, the multiple generators such that each generator is trained according to a first objective to output examples of a designated classification category;
- and
- training, through machine learning, the classifier to determine, for each generated by the multiple generators, which of the multiple generators generated the example,
- wherein the cooperative training comprises back-propagating partial derivatives of an error cost function from the classifier to the multiple generators.
A computer system comprising:
- one or more processor cores;
- and
- a memory in communication with the one or more processor cores, wherein the memory stores computer instructions that when executed by the one or more processor cores cause the one or more processor cores to cooperatively train multiple generators and a classifier by:
- training, through machine learning, the multiple generators such that each generator is trained according to a first objective to output examples of a designated classification category;
- and
- training, through machine learning, the classifier to determine, for each generated by the multiple generators, which of the multiple generators generated the example,
- wherein the cooperative training comprises back-propagating partial derivatives of an error cost function from the classifier to the multiple generators.
Dependencies are inferred from the claim text.