Continuing the motivating remarks for a prediction network, we provide a brief introduction to the use of value functions. This is intended to demonstrate why turnkey, networked microprediction isn't limited to the prediction of exogenous measured quantities, but can potentially improve applications that lean on control theory or reinforcement learning.
Artificial Intelligence is very often thinly veiled repeated prediction. It isn't prediction in the sense of predicting extant instrumented quantities, but it is repeated prediction nonetheless.
In many approaches to repeated decision making, whether it involves playing chess or navigating a room, the mathematical notion of a value function can provide a shortcut.
Recognizing that your business can be improved through the use of control theory is a first step. The second step is recognizing that applied mathematical techniques such as this can be monitored, improved or even driven from scratch by microprediction. By tapping the prediction network you can take out a large part of the expense, time and pain.