A* Search

A* Search finds a path by looking at distance already traveled plus a guess of what is left.

What it does

A* Search finds a path by looking at distance already traveled plus a guess of what is left.

Heuristic: A smart guess • Pathfinding: Finding the way

How it works

• Start at the first spot.
• Score nearby spots using cost so far and estimated distance left.
• Keep choosing promising spots until the goal is reached.

Uses and applications

• A game can move a character around walls.
• A map app can find a route.
• A warehouse robot can plan a path through shelves.

In short

A* is a smart path finder.

Related methods

• MCTS (Tree Search): Searches by sampling outcomes instead of following a cost estimate.
• Simplex Algorithm: Optimizes a goal under constraints rather than finding a path.
• AutoGPT (Agents): Plans a sequence of steps toward a goal with a language model.

Back to Reinforcement learning, agents and planning · All topics

The interactive demonstration and the Python walkthrough for this topic need JavaScript.