Contextual long-term survival optimization for content management system content selection

A non-transitory machine-readable storage medium that provides instructions that, when executed by a processor, cause the processor to:

• generate a first output using a first machine learning-based model based on context data and choice data of a first training data item;
• generate a second output using a second machine learning-based model, wherein the second output comprises an output sampled from a distribution generated based on the first output and reward data of the first training data item;
• generate a loss function output using a loss function based on the second output and the reward data of the first training data item;
• and
• update the first machine learning-based model based on the loss function output.

A system, comprising:

• A processor;
• and
• memory coupled to the processor, wherein the memory comprises instructions that cause the processor to:
• generate a first output using a first machine learning-based model based on context data and choice data of a first training data item;
• generate a second output using a second machine learning-based model, wherein the second output comprises an output sampled from a distribution generated based on the first output and reward data of the first training data item;
• generate a loss function output using a loss function based on the second output and the reward data of the first training data item;
• and
• update the first machine learning-based model based on the loss function output.