dignea  1.0.0
Diverse Instance Generator with Novelty Search and Evolutionary Algorithms
EasyInstances Class Reference

Easy Instance fitness computation. This class extends InstanceFitness. More...

#include <EasyInstances.h>

Inheritance diagram for EasyInstances:
InstanceFitness

Public Member Functions

float compute (const vector< float > &solutions) const override
 Computes the performance fitness of an instance based on the performance of different solvers. This method computes the fitness as the different between the performance of the target solver minus the maximum performance of the others. More...
 

Detailed Description

Easy Instance fitness computation. This class extends InstanceFitness.

Member Function Documentation

◆ compute()

float EasyInstances::compute ( const vector< float > &  solutions) const
overridevirtual

Computes the performance fitness of an instance based on the performance of different solvers. This method computes the fitness as the different between the performance of the target solver minus the maximum performance of the others.

solutions[0] - max(solutions[1..])

Parameters
solutions
Returns
float

Computes the fitness of an individual in the EIG. This particular formulation is based on the publicacion of Coello 2019 about Knapsack Instance Generation using Heuristics

Parameters
solutions
Returns

Implements InstanceFitness.


The documentation for this class was generated from the following files: