EvolutionaryConfigsGenAlg

This class is a child of ConfigsGenAlg. It uses an evolutionary algorithm to optimize the generation of group configurations. Note: still in development

Constructor and Attributes

Constructor

EvolutionaryConfigsGenAlg(numberOfConfigChoices, minNumberOfPlayersPerGroup, maxNumberOfPlayersPerGroup, preferredNumberOfPlayersPerGroup, fitnessWeights, qualityWeights: PlayerState, regAlg: RegressionAlg, numMutations: int, numFitSurvivors: int, probOfMutation: float)

Attributes

Name: expected type	Default value	Description
regAlg: RegressionAlg	-	The regression algorithm used to predict new states from the existing ones
numberOfConfigChoices: int	100	The number of configurations to generate, from where to choose the best one.
fitnessWeights: PlayerCharacteristics	PlayerCharacteristics(ability=0.5, engagement=0.5)	Weights to consider when calculating the quality of predicted group states (same function as qualityWeights in other classes)
numMutations: int	4	The number of allowed mutations whenever the mutation operator is selected
probOfMutation: float	0.1	The probability of selecting the mutation operator
numFitSurvivors: int	⌈numberOfConfigChoices/2⌉	The number of allowed survivors between iterations

(No more additional attributes besides the ones from the base class ConfigsGenAlg )

Methods

initPopulation(): void

Description

Internal method used to initialize the population (create random group configurations and profiles).

fitnessSort(): void

Description

Internal method used to sort the individuals of the population (group configurations and their profiles).

(No more additional methods besides the ones from the base class ConfigsGenAlg )