Integrated machine learning training

A computer-implemented method to manage machine learning models based on independently hosted customer networks, the method comprising:

• receiving a request 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 that simulates an independently hosted customer network;
• in response to the received request, selecting a first simulated, hosted customer network from a set of simulated, hosted customer networks using the received request as selection criteria;
• subsequent to the received request, obtaining training data via an interface from one or more agents associated with the simulated, hosted customer network, wherein the training data is embodied as state information of the independently hosted customer network, action information of the independently hosted customer network, reward information of the simulated, hosted customer network, and observation information of the simulated, hosted customer network;
• processing the training data in accordance with a reinforcement learning model to form the trained machine learning model;
• and
• providing the trained machine learning model in response to the request.

A computer-implemented method to manage machine learning models based on simulated, hosted customer networks, the computer-implemented method comprising:

• receiving a request to provide a trained machine learning model that is trained in accordance with a reinforcement learning model on training data from one or more identified simulated, hosted customer networks;
• in response to the received request, 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 request, causing a simulation of one or more independently hosted customer networks;
• obtaining training data via an interface from one or more agents associated with individual simulated, hosted customer networks, wherein the training data is embodied as information for implementation of a reinforcement learning machine learning algorithm;
• processing the training data in accordance with a reinforcement learning model to form the trained machine learning model without access to the one or more hosted customer networks;
• and
• providing the trained machine learning model in response to the request.