Implementation of the UCB algorithm for multiarmed bandits. More...

#include <UCB.hpp>

Public Member Functions
size_t	get_choice ()

size_t	choose ()
	Choose using the current state.

void	reward (double)
	Provide reward for the most recent choice.

Private Attributes
bool	choose_next
	Belt-and braces: warn if choose/reward happens in the wrong order.

size_t	choice
	Store the last choice made.

Static Private Attributes
static boost::random::mt19937	random_generator
	Random source.

Friends
ostream &	operator<< (ostream &, const UCB &)

Detailed Description

Implementation of the UCB algorithm for multiarmed bandits.

It's easy enough to find a description of this algorithm. The implementation here is pretty much straight from my book.

Definition at line 50 of file UCB.hpp.

Constructor & Destructor Documentation

◆ UCB()

UCB::UCB ( )

inline

Definition at line 70 of file UCB.hpp.

70{};

Member Function Documentation

◆ choose()

size_t UCB::choose ( )

Choose using the current state.

Definition at line 42 of file UCB.cpp.

                   {
    if (!choose_next) {
        cerr << "STOP IT!! EXP3 should be receiving reward..." << endl;
    }
    choose_next = false;
    
    return choice;
}

◆ get_choice()

size_t UCB::get_choice ( )

Definition at line 38 of file UCB.cpp.

                       {
    return choice;
}

◆ reward()

void UCB::reward ( double r )

Provide reward for the most recent choice.

Definition at line 51 of file UCB.cpp.

                         {
    if (choose_next) {
        cerr << "STOP IT!! EXP3 should be choosing..." << endl;
    }
    choose_next = true;
    
}

Friends And Related Symbol Documentation

◆ operator<<

ostream & operator<<	(	ostream &	out,
		const UCB &	ucb )

friend

Definition at line 59 of file UCB.cpp.

                                                  {
    
    return out;
}

Member Data Documentation

◆ choice

size_t UCB::choice

private

Store the last choice made.

Definition at line 68 of file UCB.hpp.

◆ choose_next

bool UCB::choose_next

private

Belt-and braces: warn if choose/reward happens in the wrong order.

Definition at line 64 of file UCB.hpp.

◆ random_generator

boost::random::mt19937 UCB::random_generator

staticprivate

Random source.

Underlying random number generator for boost::random::discrete_distribution.

Definition at line 58 of file UCB.hpp.

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

/Users/sbh11/Desktop/connection-prover/c++/connect++/source/bandits/UCB.hpp
/Users/sbh11/Desktop/connection-prover/c++/connect++/source/bandits/UCB.cpp

Public Member Functions

Private Attributes

Static Private Attributes

Friends