← Back to Gallery

Optimization Algorithms

Explore 10 interactive visualizations of optimization algorithms that search for optimal solutions in complex problem spaces. Watch algorithms navigate fitness landscapes, escape local optima, and converge to solutions through evolution, swarm intelligence, and mathematical techniques.

01

Gradient Descent

Watch a ball roll down a fitness landscape following the steepest slope. Adjust learning rates to see how step size affects convergence speed and stability.

02

Simulated Annealing

Temperature-based probabilistic search that accepts worse solutions early to escape local optima. See it solve the Traveling Salesman Problem as temperature cools.

03

Hill Climbing

Greedy local search that always moves uphill. Watch multiple attempts get stuck in local optima, demonstrating the fundamental limitation of greedy algorithms.

04

Genetic Algorithm (TSP)

Evolution solving the Traveling Salesman Problem through selection, crossover, and mutation. Watch populations evolve from random routes to near-optimal solutions.

05

Particle Swarm Optimization

Swarm of particles searching a landscape, balancing personal best positions with global discoveries. See emergent intelligence from simple local rules.

06

Ant Colony Optimization

Virtual ants deposit pheromone trails while exploring paths. Stronger trails attract more ants, creating positive feedback that discovers shortest routes.

07

Differential Evolution

Population-based algorithm for continuous optimization using vector differences. Watch the population contract around optimal solutions through mutation and selection.

08

Bayesian Optimization

Sample-efficient search using a surrogate model (Gaussian Process) to predict promising areas. Balances exploration vs exploitation intelligently.

09

Convex vs Non-Convex

Compare optimization on different landscape types. Convex landscapes guarantee global optima, while non-convex landscapes have many local minima traps.

10

Multi-Objective Optimization

Optimize multiple conflicting objectives simultaneously. Discover the Pareto frontier where improving one objective requires sacrificing another.