rl_logic.ValueEstimator Class Reference

Class for assigning current estimated value for a given action and provides method for returning this value. More...

Collaboration diagram for rl_logic.ValueEstimator:

Public Member Functions
def	__init__ (self, est_method_id="sample_average")
	Constructor. More...
def	estimate_value (self)
	Main method for estimation value calculation. More...
def	estimate_value_by_sample_average (self, action_id, reward)
	Estimate value by using a simple "sample average" method. More...
def	delete_action_id (self, action_id)
	Delete an action_id from the current actions list. More...

Public Attributes
	actions
	Store current action ids and their current estimated value and step: {action_id: [est_value, step_count]}. More...
	estimate_value
	A reference to the estimation method chosen by the est_method_id. More...

Detailed Description

Class for assigning current estimated value for a given action and provides method for returning this value.

Constructor & Destructor Documentation

def rl_logic.ValueEstimator.__init__	(		self,
			est_method_id = "sample_average"
	)

Constructor.

Parameters

self	The object pointer.
est_method_id	Default calculation method of the estimation value

Member Function Documentation

def rl_logic.ValueEstimator.delete_action_id	(		self,
			action_id
	)

Delete an action_id from the current actions list.

Parameters

self	The object pointer.
action_id	ID of the action being deleted.

def rl_logic.ValueEstimator.estimate_value

(

self

)

Main method for estimation value calculation.

It is being overridden in the constructor, depending on the chosen estimation method ID. Input: action_id - some action identifier; reward - value of the assigned reward. Output: current estimated value.

Parameters

self	The object pointer.

def rl_logic.ValueEstimator.estimate_value_by_sample_average	(		self,
			action_id,
			reward
	)

Estimate value by using a simple "sample average" method.

Reference to the method can be found in R.Sutton's book: Reinforcement Learning: An Introduction.

Parameters

self	The object pointer.
action_id	ID of the action having been chosen.
reward	Reward value received on the corresponding action ID.

Returns

Estimated value in float().

Member Data Documentation

rl_logic.ValueEstimator.actions

Store current action ids and their current estimated value and step: {action_id: [est_value, step_count]}.

rl_logic.ValueEstimator.estimate_value

A reference to the estimation method chosen by the est_method_id.

---

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

• rl_logic/__init__.py