bayesopt::DiscreteModel Class Reference

Bayesian optimization for functions in discrete spaces. More...

#include <bayesopt.hpp>

Inheritance diagram for bayesopt::DiscreteModel:

Collaboration diagram for bayesopt::DiscreteModel:

Public Member Functions
	DiscreteModel (const vecOfvec &validSet, Parameters params)
	Constructor for real-valued discrete data. More...
	DiscreteModel (const vectori &categories, Parameters params)
	Constructor for categorical data. More...
virtual	~DiscreteModel ()
	Default destructor.
Public Member Functions inherited from bayesopt::BayesOptBase
	BayesOptBase (size_t dim, Parameters params)
	Constructor. More...
virtual	~BayesOptBase ()
	Default destructor.
virtual double	evaluateSample (const vectord &query)=0
	Function that defines the actual function to be optimized. More...
virtual bool	checkReachability (const vectord &query)
	This function checks if the query is valid or not. More...
void	optimize (vectord &bestPoint)
	Execute the optimization process of the function defined in evaluateSample. More...
void	stepOptimization ()
	Execute ONE step the optimization process of the function defined in evaluateSample.
void	initializeOptimization ()
	Initialize the optimization process. More...
vectord	getFinalResult ()
	Once the optimization has been perfomed, return the optimal point. More...
void	saveOptimization (BOptState &state)
	Saves the current state of the optimization process into a state class. More...
void	restoreOptimization (BOptState state)
	Restores the optimization process of a previous execution.
ProbabilityDistribution *	getPrediction (const vectord &query)
const Dataset *	getData ()
Parameters *	getParameters ()
double	getValueAtMinimum ()
size_t	getCurrentIter ()
double	evaluateCriteria (const vectord &query)

Protected Member Functions
vectord	samplePoint ()
	Sample a single point in the input space. More...
void	findOptimal (vectord &xOpt)
	Call the inner optimization method to find the optimal point acording to the criteria. More...
vectord	remapPoint (const vectord &x)
	Remap the point x to the original space.
void	generateInitialPoints (matrixd &xPoints)
	Selects the initial set of points to build the surrogate model. More...
Protected Member Functions inherited from bayesopt::BayesOptBase
vectord	getPointAtMinimum ()
	Get optimal point in the inner space (e.g. More...
double	evaluateSampleInternal (const vectord &query)
	Wrapper for the target function adding any preprocessing or constraint. More...
void	plotStepData (size_t iteration, const vectord &xNext, double yNext)
	Print data for every step according to the verbose level. More...
void	saveInitialSamples (matrixd xPoints)
	Eases the process of saving a state during initial samples.
void	saveResponse (double yPoint, bool clear)

Private Member Functions
	DiscreteModel ()
	Default constructor forbidden.

Private Attributes
vecOfvec	mInputSet
	List of input points.

Additional Inherited Members
Protected Attributes inherited from bayesopt::BayesOptBase
Parameters	mParameters
	Configuration parameters.
size_t	mDims
	Number of dimensions.
size_t	mCurrentIter
	Current iteration number.
boost::mt19937	mEngine
	Random number generator.

Detailed Description

Bayesian optimization for functions in discrete spaces.

The discrete space can be created in two ways depending on the constructor used:

1. A set of discrete points (real vectors) in a real space.
2. A set of categories with different values for each category.

The kind of models used in this library are more suitable for problems of the first point. However, it can also be used with problems in the second category. In that case, we recommend to use the Hamming kernel function.

Usage:

DiscreteModel opt(validSetOfPoints,params);
\\ or
DiscreteModel opt(categories,params);

Optimization can be run in batch mode calling

opt.optimize(result);

or step by step.

opt.initiliazeOptimization();
\\...
opt.stepOptimization();
\\...
result getFinalResult();

See also

BayesOptBase about how to run the optimization

HammingKernel for categorical data.

Definition at line 166 of file bayesopt.hpp.

Constructor & Destructor Documentation

DiscreteModel() [1/2]

bayesopt::DiscreteModel::DiscreteModel	(	const vecOfvec &	validSet,
		Parameters	params
	)

Constructor for real-valued discrete data.

Parameters

validSet	Set of potential inputs
params	set of parameters (see parameters.h)

Definition at line 42 of file bayesoptdisc.cpp.

References bayesopt::BayesOptBase::mDims, and mInputSet.

DiscreteModel() [2/2]

bayesopt::DiscreteModel::DiscreteModel	(	const vectori &	categories,
		Parameters	params
	)

Constructor for categorical data.

Parameters

number	of categories per dimension
params	set of parameters (see parameters.h)

Definition at line 49 of file bayesoptdisc.cpp.

References bayesopt::BayesOptBase::mDims, and mInputSet.

Member Function Documentation

findOptimal()

void bayesopt::DiscreteModel::findOptimal ( vectord &  xOpt )

protectedvirtual

Call the inner optimization method to find the optimal point acording to the criteria.

Parameters

xOpt	optimal point

Implements bayesopt::BayesOptBase.

Definition at line 70 of file bayesoptdisc.cpp.

References mInputSet.

generateInitialPoints()

void bayesopt::DiscreteModel::generateInitialPoints ( matrixd &  xPoints )

protectedvirtual

Selects the initial set of points to build the surrogate model.

Implements bayesopt::BayesOptBase.

Definition at line 98 of file bayesoptdisc.cpp.

References bayesopt::BayesOptBase::mEngine, mInputSet, and bayesopt::utils::randomPerms().

samplePoint()

vectord bayesopt::DiscreteModel::samplePoint ( )

protectedvirtual

Sample a single point in the input space.

Used for epsilon greedy exploration.

Implements bayesopt::BayesOptBase.

Definition at line 64 of file bayesoptdisc.cpp.

References bayesopt::BayesOptBase::mEngine.

---

The documentation for this class was generated from the following files:

• bayesopt.hpp
• bayesoptdisc.cpp