Tutorials

Hands-on, end-to-end walkthroughs of real tasks. Each one is self-contained and runnable top to bottom. If you are new, start with Getting started first.

• Tuning a model's hyperparameters — use PSOSearchCV as a drop-in alternative to GridSearchCV, tune an SVM on a real dataset, and read the results.
• Visualizing and animating a swarm — record a run, plot convergence, compare variants, and animate the swarm over the objective landscape.
• Constrained optimization — solve a problem with inequality constraints using the penalty method and verify the optimum.
• Multi-objective optimization — find and visualize a Pareto front of conflicting objectives with MOPSO, and measure it with hypervolume.
• Integer and mixed optimization — solve a 0/1 knapsack (and verify the global optimum) and optimize real/integer/binary variables together.
• Gradient-free neural-network training — train a small PyTorch model by optimizing its weights on a non-differentiable metric (accuracy) with PSO.

Each tutorial only needs turboswarm plus the extras it calls out (e.g. pip install turboswarm[sklearn] for the first one).