reinforcement learning and hill-climbing optimization

Local maxima and plateaus

The loop design mirrors local search: agents exploit a current objective, while human judgment supplies the exploratory move to a new optimum.

1 source videos