Tutorials

This is a collection of tutorials on PyGMO’s basics:

• Adding new problems and algorithms to PyGMO
  • Adding a new optimization problem
    • Simple single objective problem
    • Maximization problem
    • Multi-objective problem
  • Adding a new stochastic optimization problem
  • Adding a new algorithm
• Meta-problems
  • Rotated and Shifted problems
  • The Noisy meta-problem
  • Decomposition
• Multi Objective Optimization with PyGMO
  • Basic Multi-Objective Functionalities
  • PADE: a parallel MOEA-D algorithm
  • Using NSGA-II, SPEA2 and NS-PSO
• Hypervolumes with PyGMO
  • Getting started with hypervolumes
    • Hypervolume interface and construction
    • Computing the hypervolume indicator and hypervolume contributions
  • Advanced hypervolume computation and analysis
    • Computing hypervolume using a specific algorithm
    • Runtime analysis of the hypervolume computation
    • Algorithm comparison
      • Short summary of algorithms
    • Pushing efficiency further
      • Bonus
  • Approximating the hypervolume
    • FPRAS
      • Running time
    • Approximation of the least contributor
  • Migration based on hypervolume contribution
    • How does the migration work ?
      • Hypervolume-based selection policy
      • Hypervolume-based replacement policy
  • Using hypervolumes in your own algorithm
• Dealing with constraints in PyGMO
  • Death penalty
    • Method
    • Application
  • Multi-objective transformation
    • Method
    • Application
  • Self-adaptive penalty
    • Method
    • Application
  • Co-evolution penalty method
    • Method
    • Application
  • Immune system method for constraint handling
    • Method
    • Application
  • Repair methods
    • Method
    • Application
• Racing in PyGMO
  • Racing the individuals in stochastic optimization problems
    • Background
    • Racing with PyGMO
    • Why racing: A comparative study
      • Varying noise level
      • Varying initial population sizes
      • Varying evaluation budget
  • Racing algorithms
    • Using race_algo to race algorithms
    • Example: Identifying good configurations of an algorithm
      • Why not choosing a single best winner?
  • A PSO algorithm using racing
• Traveling Salesman Problems with PyGMO
  • Introduction
    • TSP as integer linear programs (FULL encoding)
    • TSP encoded using Random Keys (RANDOMKEYS encoding)
    • TSP encoded using the city sequence (CITIES encoding)
  • Instantiating a TSP problem
    • From weights
    • From TSPLIB XML
  • Solving a TSP
• Landscape Analysis using Dr. PyGMO
  • Analysis of the Fitness Landscape
    • Instantiate Analysis Module
    • Tests on fitness functions
  • Analysis of the Constraints Landscape
    • Instantiate Analysis Module
    • Tests on constraints functions