BayesOpt
bayesopt::ProbabilityDistribution Class Referenceabstract
+ Inheritance diagram for bayesopt::ProbabilityDistribution:

Public Member Functions

 ProbabilityDistribution (randEngine &eng)
 
virtual double pdf (double x)=0
 Probability density function. More...
 
virtual double negativeExpectedImprovement (double min, size_t g)=0
 Expected Improvement algorithm for minimization. More...
 
virtual double lowerConfidenceBound (double beta=1)=0
 Lower confindence bound. More...
 
virtual double negativeProbabilityOfImprovement (double yMin, double epsilon)=0
 Probability of improvement algorithm for minimization. More...
 
virtual double sample_query ()=0
 Sample outcome acording to the marginal distribution at the query point. More...
 
virtual double getMean ()=0
 
virtual double getStd ()=0
 

Protected Attributes

randEngine & mtRandom
 

Detailed Description

Definition at line 35 of file prob_distribution.hpp.

Member Function Documentation

◆ lowerConfidenceBound()

virtual double bayesopt::ProbabilityDistribution::lowerConfidenceBound ( double  beta = 1)
pure virtual

Lower confindence bound.

Can be seen as the inverse of the Upper confidence bound

Parameters
betastd coefficient (used for annealing)
Returns
value of the lower confidence bound

Implemented in bayesopt::StudentTDistribution, and bayesopt::GaussianDistribution.

◆ negativeExpectedImprovement()

virtual double bayesopt::ProbabilityDistribution::negativeExpectedImprovement ( double  min,
size_t  g 
)
pure virtual

Expected Improvement algorithm for minimization.

Parameters
minminimum value found so far
gexponent (used for annealing)
Returns
negative value of the expected improvement

Implemented in bayesopt::StudentTDistribution, and bayesopt::GaussianDistribution.

◆ negativeProbabilityOfImprovement()

virtual double bayesopt::ProbabilityDistribution::negativeProbabilityOfImprovement ( double  yMin,
double  epsilon 
)
pure virtual

Probability of improvement algorithm for minimization.

Parameters
minminimum value found so far
epsilonminimum improvement margin
Returns
negative value of the probability of improvement

Implemented in bayesopt::StudentTDistribution, and bayesopt::GaussianDistribution.

◆ pdf()

virtual double bayesopt::ProbabilityDistribution::pdf ( double  x)
pure virtual

Probability density function.

Parameters
xquery point
Returns
probability

Implemented in bayesopt::StudentTDistribution, and bayesopt::GaussianDistribution.

◆ sample_query()

virtual double bayesopt::ProbabilityDistribution::sample_query ( )
pure virtual

Sample outcome acording to the marginal distribution at the query point.

Parameters
engboost.random engine
Returns
outcome

Implemented in bayesopt::StudentTDistribution, and bayesopt::GaussianDistribution.


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