Package | Description |
---|---|
org.processmining.plugins.etm.parameters | |
org.uncommonseditedbyjoosbuijs.watchmaker.framework |
This package provides a framework for evolutionary computation.
|
org.uncommonseditedbyjoosbuijs.watchmaker.framework.interactive |
Classes for implementing interactive evolutionary algorithms.
|
org.uncommonseditedbyjoosbuijs.watchmaker.framework.islands |
An implementation of island model evolution.
|
org.uncommonseditedbyjoosbuijs.watchmaker.framework.selection |
Various selection strategies for use with evolutionary algorithms.
|
Modifier and Type | Field and Description |
---|---|
protected SelectionStrategy<java.lang.Object> |
ETMParamAbstract.selectionStrategy |
Modifier and Type | Method and Description |
---|---|
SelectionStrategy<java.lang.Object> |
ETMParamAbstract.getSelectionStrategy() |
Modifier and Type | Method and Description |
---|---|
void |
ETMParamAbstract.setSelectionStrategy(SelectionStrategy<java.lang.Object> selectionStrategy) |
Modifier and Type | Field and Description |
---|---|
protected SelectionStrategy<? super T> |
GenerationalEvolutionEngine.selectionStrategy |
Constructor and Description |
---|
GenerationalEvolutionEngine(CandidateFactory<T> candidateFactory,
EvolutionaryOperator<T> evolutionScheme,
FitnessEvaluator<? super T> fitnessEvaluator,
SelectionStrategy<? super T> selectionStrategy,
java.util.Random rng)
Creates a new evolution engine by specifying the various components
required by a generational evolutionary algorithm.
|
SteadyStateEvolutionEngine(CandidateFactory<T> candidateFactory,
EvolutionaryOperator<T> evolutionScheme,
FitnessEvaluator<? super T> fitnessEvaluator,
SelectionStrategy<? super T> selectionStrategy,
int selectionSize,
boolean forceSingleCandidateUpdate,
java.util.Random rng)
Create a steady-state evolution strategy in which one or more (usually
just one) evolved offspring replace randomly-chosen individuals.
|
Modifier and Type | Class and Description |
---|---|
class |
InteractiveSelection<T>
Special selection strategy used for interactive evolutionary algorithms.
|
Constructor and Description |
---|
IslandEvolution(int islandCount,
Migration migration,
CandidateFactory<T> candidateFactory,
EvolutionaryOperator<T> evolutionScheme,
FitnessEvaluator<? super T> fitnessEvaluator,
SelectionStrategy<? super T> selectionStrategy,
java.util.Random rng)
Create an island system with the specified number of
identically-configured islands.
|
Modifier and Type | Class and Description |
---|---|
class |
RankSelection
A selection strategy that is similar to fitness-proportionate selection
except that is uses relative fitness rather than absolute fitness in order to
determine the probability of selection for a given individual (i.e.
|
class |
RouletteWheelSelection
Implements selection of n candidates from a population by selecting
n candidates at random where the probability of each candidate getting
selected is proportional to its fitness score.
|
class |
SigmaScaling
An alternative to straightforward fitness-proportionate selection such as that offered
by
RouletteWheelSelection and StochasticUniversalSampling . |
class |
StochasticUniversalSampling
An alternative to
RouletteWheelSelection
as a fitness-proportionate selection strategy. |
class |
TournamentSelection
Selection strategy that picks a pair of candidates at random and then
selects the fitter of the two candidates with probability p, where p
is the configured selection probability (therefore the probability of
the less fit candidate being selected is 1 - p).
|
class |
TruncationSelection
Implements selection of n candidates from a population by simply
selecting the n candidates with the highest fitness scores (the
rest are discarded).
|
Constructor and Description |
---|
RankSelection(SelectionStrategy<java.lang.Object> delegate)
Creates a rank-based selector with a linear mapping function and
configurable delegate for performing the proportionate selection.
|
SigmaScaling(SelectionStrategy<java.lang.Object> delegate)
Creates a sigma-scaled selection strategy that delegates to the specified selection
strategy after adjusting individual fitness scores using sigma-scaling.
|