A municipal transit agency trains a traffic-signal controller that chooses a phase plan at each intersection. After every cycle the controller receives a numeric score based on leftover queue length and pedestrian wait time. Engineers have no labeled dataset of correct phase plans. Which learning approach is the agency using?
Select an answer to reveal the explanation.
Short Explanation
Think of it like teaching a dog with treats instead of flashcards. The controller tries a phase plan, the intersection scores leftover queues and wait time, and the next choice gets a little smarter — that is a reward loop, not a labeled answer key. Supervised learning would need the "right" plans already written down, and unsupervised learning just looks for clumps in the data with no score at all.
Full Explanation
Reinforcement learning trains an agent through trial and feedback from an environment, using a reward or penalty instead of labeled correct outputs. The signal controller picks a phase plan and is scored on queues and wait time, with no labeled set of right plans, which matches that pattern. Supervised learning needs labeled input-output pairs. Unsupervised learning, including association rule mining, finds structure in unlabeled data without a reward signal.