SmartEngine
1.6.0
|
GeneticTrainer parameter info More...
Public Attributes | |
float | CountTopPerformers = 0.1f |
How many of the top chromosomes are marked as top performers. The mutation standard deviation will be adjusted depending on if a new chromosome enters the top performers list. More... | |
float | CountToKeepUnchanged = 0.4f |
How many of the top chromosomes should be left unchanged during each step. More... | |
float | CountToRandomize = 0.1f |
How many of the bottom chromosomes should be completely randomized. More... | |
readonly MutationInfo | Mutation = new MutationInfo() |
Mutation info More... | |
readonly GeneSwapInfo | GeneSwap = new GeneSwapInfo() |
Gene swap info More... | |
GeneticTrainer parameter info
float SmartEngine.NeuralNetworks.GeneticTrainingInfo.CountToKeepUnchanged = 0.4f |
How many of the top chromosomes should be left unchanged during each step.
A value [0..1] will unchange that percent of the population
A negative integer (-Population Count, 0) will keep that exact number of chromosomes
float SmartEngine.NeuralNetworks.GeneticTrainingInfo.CountTopPerformers = 0.1f |
How many of the top chromosomes are marked as top performers. The mutation standard deviation will be adjusted depending on if a new chromosome enters the top performers list.
A value [0..1] equates to a percent of the population
A negative integer (-Population Count, 0) equates to that exact number of chromosomes.
float SmartEngine.NeuralNetworks.GeneticTrainingInfo.CountToRandomize = 0.1f |
How many of the bottom chromosomes should be completely randomized.
A value [0..1] will randomize that percent of the population
A negative integer (-Population Count, 0) will randomize that exact number of the population
readonly GeneSwapInfo SmartEngine.NeuralNetworks.GeneticTrainingInfo.GeneSwap = new GeneSwapInfo() |
Gene swap info
readonly MutationInfo SmartEngine.NeuralNetworks.GeneticTrainingInfo.Mutation = new MutationInfo() |
Mutation info