Search Landscape Features

Interactive Exploration of Hill Climbing Challenges

Click anywhere to place the agent and watch hill climbing in action!

Common Search Landscape Features

🏆 Global Maximum
The highest point overall - our goal!
🔺 Local Maximum
Higher than neighbors, but not the best
➖ Plateau
Flat areas with no improving moves
🪜 Shoulder
Flat areas that eventually lead upward
↔️ Ridge
Narrow paths where progress is tricky

Interactive Search Landscape

--
Position X
--
Position Y
--
Current Utility
0
Steps Taken
Ready
Algorithm Status
Click anywhere on the landscape to place the agent!
Watch how hill climbing behaves in different terrain features.

Demonstration Scenarios

Click any scenario to see hill climbing behavior in different landscape features:

Find Global Maximum

Start near the global peak and watch success!

Trapped in Local Maximum

Get stuck on a smaller peak - classic hill climbing problem!

Lost on Plateau

Wander aimlessly on flat terrain with no direction!

Shoulder Challenge

Navigate a flat shelf that leads to higher ground!

Ridge Navigation

Follow a narrow path where small wrong steps lead downward!

Place the agent by clicking on the landscape, then start the hill climbing algorithm.

Key Educational Insights

🔺 Local Maximum Problem

Hill climbing stops at the first peak it finds, even if taller peaks exist elsewhere. This is why the algorithm is called "greedy" - it takes the best immediate option without considering the bigger picture.

➖ Plateau Challenge

On flat terrain, hill climbing has no gradient to follow. All neighbors have the same utility, so the algorithm can't decide which direction leads to improvement.

🪜 Shoulder Opportunity

Shoulders are flat regions that eventually lead upward. A patient algorithm that explores multiple moves can find the hidden upward path.

↔️ Ridge Difficulty

Ridges require precise navigation. One wrong step leads downward, making progress slow and requiring careful exploration of each move.