Supervised regression learning functions. More...

#include "xcs_supervised.h"
#include "clset.h"
#include "ea.h"
#include "loss.h"
#include "pa.h"
#include "param.h"
#include "perf.h"
#include "utils.h"

Include dependency graph for xcs_supervised.c:

Functions
static int	xcs_supervised_sample (const struct Input *data, const int cnt, const bool shuffle)
	Selects a data sample for training or testing.

static void	xcs_supervised_trial (struct XCSF xcsf, const double x, const double y, const double cover)
	Executes a single XCSF trial.

double	xcs_supervised_fit (struct XCSF xcsf, const struct Input train_data, const struct Input *test_data, const bool shuffle, const int start, const int trials)
	Executes MAX_TRIALS number of XCSF learning iterations using the training data and test iterations using the test data.

void	xcs_supervised_predict (struct XCSF xcsf, const double x, double pred, const int n_samples, const double cover)
	Calculates the XCSF predictions for the provided input.

double	xcs_supervised_score (struct XCSF xcsf, const struct Input data, const double *cover)
	Calculates the XCSF error for the input data.

double	xcs_supervised_score_n (struct XCSF xcsf, const struct Input data, const int N, const double *cover)
	Calculates the XCSF error for a subsample of the input data.

Detailed Description

Supervised regression learning functions.

Author: Richard Preen rpree.nosp@m.n@gm.nosp@m.ail.c.nosp@m.om

Copyright: The Authors.

Date: 2015–2024.

Definition in file xcs_supervised.c.

Function Documentation

◆ xcs_supervised_fit()

double xcs_supervised_fit	(	struct XCSF *	xcsf,
		const struct Input *	train_data,
		const struct Input *	test_data,
		const bool	shuffle,
		const int	start,
		const int	trials
	)

Executes MAX_TRIALS number of XCSF learning iterations using the training data and test iterations using the test data.

Parameters

[in]	xcsf	The XCSF data structure.
[in]	train_data	The input data to use for training.
[in]	test_data	The input data to use for testing.
[in]	shuffle	Whether to randomise the instances during training.
[in]	start	Index to begin sampling.
[in]	trials	Number of trials to execute.

Returns: The average XCSF training error using the loss function.

Definition at line 99 of file xcs_supervised.c.

References param_set_explore(), perf_print(), Input::x, Input::x_dim, xcs_supervised_sample(), xcs_supervised_trial(), Input::y, and Input::y_dim.

Referenced by XCS::fit(), and main().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ xcs_supervised_predict()

void xcs_supervised_predict	(	struct XCSF *	xcsf,
		const double *	x,
		double *	pred,
		const int	n_samples,
		const double *	cover
	)

Calculates the XCSF predictions for the provided input.

Parameters

[in]	xcsf	The XCSF data structure.
[in]	x	The input feature variables.
[out]	pred	The calculated XCSF predictions.
[in]	n_samples	The number of instances.
[in]	cover	If cover is not NULL and the match set is empty, the prediction array will be set to this value instead of covering.

Definition at line 141 of file xcs_supervised.c.

References param_set_explore(), and xcs_supervised_trial().

Referenced by XCS::predict().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ xcs_supervised_sample()

static int xcs_supervised_sample	(	const struct Input *	data,
		const int	cnt,
		const bool	shuffle
	)

static

Selects a data sample for training or testing.

Parameters

[in]	data	The input data.
[in]	cnt	The current sequence counter.
[in]	shuffle	Whether to select the sample randomly.

Returns: The row of the data sample selected.

Definition at line 41 of file xcs_supervised.c.

References Input::n_samples, and rand_uniform_int().

Referenced by xcs_supervised_fit(), and xcs_supervised_score_n().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ xcs_supervised_score()

double xcs_supervised_score	(	struct XCSF *	xcsf,
		const struct Input *	data,
		const double *	cover
	)

Calculates the XCSF error for the input data.

Parameters

[in]	xcsf	The XCSF data structure.
[in]	data	The input data to calculate the error.
[in]	cover	If cover is not NULL and the match set is empty, the prediction array will be set to this value instead of covering.

Returns: The average XCSF error using the loss function.

Definition at line 161 of file xcs_supervised.c.

References Input::n_samples, param_set_explore(), Input::x, Input::x_dim, xcs_supervised_trial(), Input::y, and Input::y_dim.

Referenced by XCS::fit(), XCS::score(), and xcs_supervised_score_n().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ xcs_supervised_score_n()

double xcs_supervised_score_n	(	struct XCSF *	xcsf,
		const struct Input *	data,
		const int	N,
		const double *	cover
	)

Calculates the XCSF error for a subsample of the input data.

Parameters

[in]	xcsf	The XCSF data structure.
[in]	data	The input data to calculate the error.
[in]	N	The maximum number of samples to draw randomly for scoring.
[in]	cover	If cover is not NULL and the match set is empty, the prediction array will be set to this value instead of covering.

Returns: The average XCSF error using the loss function.

Definition at line 185 of file xcs_supervised.c.

References Input::n_samples, param_set_explore(), Input::x, Input::x_dim, xcs_supervised_sample(), xcs_supervised_score(), xcs_supervised_trial(), Input::y, and Input::y_dim.

Referenced by XCS::score().

Here is the call graph for this function:

Here is the caller graph for this function:

◆ xcs_supervised_trial()

static void xcs_supervised_trial	(	struct XCSF *	xcsf,
		const double *	x,
		const double *	y,
		const double *	cover
	)

static

Executes a single XCSF trial.

Parameters

[in]	xcsf	The XCSF data structure.
[in]	x	The feature variables.
[in]	y	The labelled variables.
[in]	cover	If cover is not NULL and the match set is empty, the prediction array will be set to this value instead of covering.

Definition at line 59 of file xcs_supervised.c.

References clset_free(), clset_init(), clset_kill(), clset_match(), clset_update(), ea(), and pa_build().

Referenced by xcs_supervised_fit(), xcs_supervised_predict(), xcs_supervised_score(), and xcs_supervised_score_n().

Here is the call graph for this function:

Here is the caller graph for this function:

Functions

Detailed Description

Function Documentation

◆ xcs_supervised_fit()

◆ xcs_supervised_predict()

◆ xcs_supervised_sample()

◆ xcs_supervised_score()

◆ xcs_supervised_score_n()

◆ xcs_supervised_trial()