Learning models can find a lot of useful applications in investing and gaming.
Reinforcement learning models how individuals choose an action to take. The probability of choosing a course of an action relative to other actions is based on how rapidly the person adjusts to new reality, the original probability of choosing this action, as well the reward beyond a level of aspiration.
So, as rewards increase, learning becomes faster. Also a positive surprise where outcomes exceeds expectations, learning can also be accelerated.
Adding a level of complication to this is that sometimes, aspirations are often endogenous. Aspirations tend to keep going upwards until it reaches the a level equal to the mean reward.
Learning can also be social. People learn by watching others. Models of social learning depend on now just how rewarding a behavior is, but also how popular this behavior is. So, if someone is doing something rewarding, then I die
...