US 12,118,456 B1
Integrated machine learning training
Sahika Genc, Mercer Island, WA (US); Bharathan Balaji, Seattle, WA (US); Urvashi Chowdhary, Seattle, WA (US); Leo Parker Dirac, Seattle, WA (US); Saurabh Gupta, Redmond, WA (US); Vineet Khare, Redmond, WA (US); and Sunil Mallya Kasaragod, San Francisco, CA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Nov. 21, 2018, as Appl. No. 16/198,730.
This patent is subject to a terminal disclaimer.
Int. Cl. G06N 3/08 (2023.01); G06F 30/20 (2020.01)
CPC G06N 3/08 (2013.01) [G06F 30/20 (2020.01)] 20 Claims
OG exemplary drawing
 
1. A system for implementation of a reinforcement machine learning environment, the system comprising:
a plurality of computing devices corresponding to a simulation environment for a set of simulated, hosted customer networks, wherein individual simulated, hosted customer networks include a convolutional neural network agent process for generating and transmitting training data, wherein the training data is embodied as state information of the hosted customer network, action information of the hosted customer network, reward information of the hosted customer network, and observation information of the simulated, hosted customer network; and
a non-transitory computer readable medium storing instructions that, when executed by the at least a processing device, cause the system to perform operations including:
receive a request from a customer computing device to provide a trained machine learning model that is trained in accordance with a reinforcement learning model on training data generated from an identified simulated, hosted customer network;
responsive to the request from the customer computing device, selecting a first simulated, hosted customer network from a set of simulated, hosted customer networks using the received request as selection criteria;
responsive to the received request, obtain the training data from a convolutional neural network agent associated with the first simulated, hosted customer network;
process the training data in accordance with the reinforcement learning model to form the trained machine learning model; and
provide, to the customer computing device, access to the trained machine learning model in response to the request.