opennn::GeneticAlgorithm Class Reference

Genetic-algorithm based input feature selection. More...

#include <genetic_algorithm.h>

Inheritance diagram for opennn::GeneticAlgorithm:

Public Member Functions
	GeneticAlgorithm (TrainingStrategy *training_strategy=nullptr)
	Constructs the algorithm.
Index	get_individuals_number () const
	Number of individuals in the current population.
Index	get_genes_number () const
	Number of genes per chromosome (= number of original inputs).
void	set_default ()
	Resets all hyperparameters to their default values.
Index	get_minimum_inputs_number () const override
	Minimum number of selected inputs allowed in any individual.
Index	get_maximum_inputs_number () const override
	Maximum number of selected inputs allowed in any individual.
void	set_minimum_inputs_number (const Index new_minimum)
	Sets the minimum number of selected inputs.
void	set_maximum_inputs_number (const Index)
	Sets the maximum number of selected inputs.
void	set_individuals_number (const Index new_individuals_number=4)
	Sets the population size.
void	set_initialization_method (string method)
	Sets the population initialization strategy.
void	set_mutation_rate (const float rate)
	Sets the mutation rate.
void	set_elitism_size (const Index size)
	Sets the elitism size.
InputsSelectionResults	perform_input_selection () override
	Runs the genetic algorithm.
void	from_JSON (const JsonDocument &) override
	Loads hyperparameters from a parsed JSON document.
void	to_JSON (JsonWriter &) const override
	Writes hyperparameters to a streaming JSON writer.
Public Member Functions inherited from opennn::InputsSelection
	InputsSelection (TrainingStrategy *training_strategy=nullptr)
	Constructs the selector.
virtual	~InputsSelection ()=default
	Virtual destructor.
const TrainingStrategy *	get_training_strategy () const
	Read-only access to the bound training strategy.
bool	has_training_strategy () const
	Whether a training strategy has been bound.
bool	get_display () const
	Whether progress should be printed to stdout.
void	set (TrainingStrategy *new_training_strategy)
	Re-initializes the selector by setting its training strategy.
void	set_trials_number (const Index new_trials_number)
	Sets the number of training trials per candidate (mean is used).
void	set_display (bool new_display)
	Toggles per-iteration progress printing.
void	set_validation_error_goal (const float new_validation_error_goal)
	Sets the validation error goal.
void	set_maximum_epochs (const Index new_maximum_epochs)
	Sets the maximum training epochs per evaluation.
void	set_maximum_validation_failures (const Index new_maximum_validation_failures)
	Sets the maximum number of consecutive validation-error increases tolerated.
void	set_maximum_time (const float new_maximum_time)
	Sets the maximum wall-clock selection time.
void	check () const
	Validates the bound configuration; throws if anything is missing.
string	get_name () const
	Canonical name of the selector (set by subclasses).
void	save (const filesystem::path &) const
	Saves the selector configuration to a file.
void	load (const filesystem::path &)
	Loads the selector configuration from a file.
virtual void	print () const
	Prints a human-readable summary of the selector to stdout.

Additional Inherited Members
Public Types inherited from opennn::InputsSelection
enum class	StoppingCondition {   MaximumTime , SelectionErrorGoal , MaximumInputs , MaximumEpochs ,   MaximumSelectionFailures }
	Reasons that can terminate an input-selection run. More...
Protected Attributes inherited from opennn::InputsSelection
TrainingStrategy *	training_strategy = nullptr
	Training strategy used to evaluate candidate subsets; not owned.
Index	trials_number = 1
	Number of independent training runs averaged per candidate.
bool	display = true
	Whether progress should be printed to stdout during selection.
float	validation_error_goal
	Validation error goal; selection stops when reached.
Index	maximum_epochs
	Maximum training epochs per evaluation.
Index	maximum_validation_failures = 100
	Maximum consecutive validation-error increases tolerated.
float	maximum_time
	Maximum wall-clock selection time in seconds.
string	name
	Canonical name of the selector (set by subclasses).

Detailed Description

Genetic-algorithm based input feature selection.

Encodes each candidate input subset as a binary chromosome (one gene per input variable), evolves a population of chromosomes through selection, crossover and mutation, and returns the subset that achieved the lowest validation error.

Constructor & Destructor Documentation

GeneticAlgorithm()

opennn::GeneticAlgorithm::GeneticAlgorithm

(

TrainingStrategy *

training_strategy = nullptr

)

Constructs the algorithm.

Parameters

training_strategy

Training strategy used to evaluate candidates.

Member Function Documentation

from_JSON()

void opennn::GeneticAlgorithm::from_JSON ( const JsonDocument & )

overridevirtual

Loads hyperparameters from a parsed JSON document.

Implements opennn::InputsSelection.

get_genes_number()

Index opennn::GeneticAlgorithm::get_genes_number ( ) const

inline

Number of genes per chromosome (= number of original inputs).

get_individuals_number()

Index opennn::GeneticAlgorithm::get_individuals_number ( ) const

inline

Number of individuals in the current population.

get_maximum_inputs_number()

Index opennn::GeneticAlgorithm::get_maximum_inputs_number ( ) const

inlineoverridevirtual

Maximum number of selected inputs allowed in any individual.

Implements opennn::InputsSelection.

get_minimum_inputs_number()

Index opennn::GeneticAlgorithm::get_minimum_inputs_number ( ) const

inlineoverridevirtual

Minimum number of selected inputs allowed in any individual.

Implements opennn::InputsSelection.

perform_input_selection()

InputsSelectionResults opennn::GeneticAlgorithm::perform_input_selection ( )

overridevirtual

Runs the genetic algorithm.

Returns

Best-of-run input subset and supporting statistics.

Implements opennn::InputsSelection.

set_default()

void opennn::GeneticAlgorithm::set_default

(

)

Resets all hyperparameters to their default values.

set_elitism_size()

void opennn::GeneticAlgorithm::set_elitism_size ( const Index size )

inline

Sets the elitism size.

Parameters

size	Number of best individuals carried unchanged to the next generation, clamped to [0, individuals_number].

set_individuals_number()

void opennn::GeneticAlgorithm::set_individuals_number

(

const Index

new_individuals_number = 4

)

Sets the population size.

Parameters

new_individuals_number

Number of individuals to maintain.

set_initialization_method()

void opennn::GeneticAlgorithm::set_initialization_method ( string method )

inline

Sets the population initialization strategy.

Parameters

method

Strategy name ("Random" or "Correlations").

set_maximum_inputs_number()

void opennn::GeneticAlgorithm::set_maximum_inputs_number

(

const Index

)

Sets the maximum number of selected inputs.

Receives the new upper bound on the number of selected inputs.

set_minimum_inputs_number()

void opennn::GeneticAlgorithm::set_minimum_inputs_number ( const Index new_minimum )

inline

Sets the minimum number of selected inputs.

Parameters

new_minimum

New lower bound on the number of selected inputs.

set_mutation_rate()

void opennn::GeneticAlgorithm::set_mutation_rate ( const float rate )

inline

Sets the mutation rate.

Parameters

rate	Per-gene flip probability, clamped to [0, 1].

to_JSON()

void opennn::GeneticAlgorithm::to_JSON ( JsonWriter & ) const

overridevirtual

Writes hyperparameters to a streaming JSON writer.

Implements opennn::InputsSelection.