Distribution Class for PBIL algorithm (Population-Based Incremental Learning, Baluja and Caruana 96) More...

Inheritance diagram for eoPBILDistrib< EOT >:

Public Member Functions
	eoPBILDistrib (unsigned _genomeSize)
	Ctor with size of genomes, and update parameters.
virtual void	operator() (EOT &_eo)
	the randomizer of indis
unsigned	Size ()
	Accessor to the genome size.
virtual void	printOn (std::ostream &os) const
	printing...
virtual void	readFrom (std::istream &is)
	reading...
unsigned int	size ()
virtual std::string	className () const
	className: Mandatory because of eoCombinedInit.
Private Attributes
unsigned	genomeSize

Detailed Description

Distribution Class for PBIL algorithm (Population-Based Incremental Learning, Baluja and Caruana 96)

It encodes a univariate distribution on the space of bitstrings, i.e. one probability for each bit to be one

It is an eoValueParam<std::vector<double> > : the std::vector<double> stores the probabilities that each bit is 1

It is still pure virtual, as the update method needs to be specified

Definition at line 44 of file eoPBILDistrib.h.

Member Function Documentation

template<class EOT>

virtual void eoPBILDistrib< EOT >::printOn ( std::ostream & os ) const [inline, virtual]

printing...

Implements eoPrintable.

Definition at line 68 of file eoPBILDistrib.h.

template<class EOT>

virtual void eoPBILDistrib< EOT >::readFrom ( std::istream & is ) [inline, virtual]

reading...

Definition at line 76 of file eoPBILDistrib.h.

template<class EOT>

virtual std::string eoPBILDistrib< EOT >::className ( void ) const [inline, virtual]

className: Mandatory because of eoCombinedInit.

SHould be pure virtual, but then we should go over the whole code to write the method for all derived classes ... MS 16/7/04

Reimplemented from eoInit< EOT >.

Definition at line 94 of file eoPBILDistrib.h.

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