Algorithms · Visualization
Evolutionary Lab
Algorithms you can actually watch.
A browser-based lab comparing genetic algorithms, particle swarm optimization and differential evolution on benchmark functions.
Public interactive demo
01 / The problem
The problem
Make optimization searches easier to understand by observing populations and comparing their convergence.
02 / Context & contribution
Context & contribution
A project published on my GitHub. Open source and contribution history make the implementation inspectable; the features below are documented in its README.
03 / Engineering choices
Engineering choices
Entirely in the browser
JavaScript runs the algorithms without a backend, making experimentation easier to access.
Visual comparison
Plotly displays 3D surfaces, populations and convergence curves for GA, PSO and DE.
Explorable parameters
Controls let visitors adjust hyperparameters and experiment with functions in Portuguese or English.
Project limitations
An exploration tool. Results depend on the function, parameters and randomness; a single run does not establish that one algorithm is better.
Explore the evidence
The demo and public code allow experiments to be reproduced. The README documents algorithm operators and interface options.
Connecting mathematical foundations to an interface that invites experimentation.
Talk about this project ↗