eshyperneat

package
v0.0.0-...-8d87d8f Latest Latest
Warning

This package is not in the latest version of its module.

Go to latest
Published: Mar 20, 2021 License: MIT Imports: 5 Imported by: 0

Documentation

Index

Constants

This section is empty.

Variables

This section is empty.

Functions

This section is empty.

Types

type ESHyperNEATContext

type ESHyperNEATContext struct {
	// The included HyperNEAT context
	*hyperneat.HyperNEATContext

	// InitialDepth defines the initial ES-HyperNEAT sample resolution.
	InitialDepth int
	// Maximal ES-HyperNEAT sample resolution if the variance is still higher than the given division threshold
	MaximalDepth int

	// DivisionThreshold defines the division threshold. If the variance in a region is greater than this value, after
	// the initial resolution is reached, ES-HyperNEAT will sample down further (values greater than 1.0 will disable
	// this feature). Note that sampling at really high resolutions can become computationally expensive.
	DivisionThreshold float64
	// VarianceThreshold defines the variance threshold for the initial sampling. The bigger this value the less new
	// connections will be added directly and the more chances that the new collection will be included in bands
	// (see BandingThreshold)
	VarianceThreshold float64
	// BandingThreshold defines the threshold that determines when points are regarded to be in a band. If the point
	// is in the band then no new connection will be added and as result no new hidden node will be introduced.
	// The bigger this value the less connections/hidden nodes will be added, i.e. wide bands approximation.
	BandingThreshold float64

	// ESIterations defines how many times ES-HyperNEAT should iteratively discover new hidden nodes.
	ESIterations int
}

ES-HyperNEAT execution context

func Load

func Load(r io.Reader) (*ESHyperNEATContext, error)

Loads ESHyperNEAT context options from provided reader

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL