lookupActionSelection: Converts a name into an action selection function

In nproellochs/ReinforcementLearning: Model-Free Reinforcement Learning

Description

Input is a name for the action selection, output is the corresponding function object.

Usage

lookupActionSelection(type)

Arguments

type	A string denoting the type of action selection. Allowed values are epsilon-greedy or random.

Value

Function that implements the specific learning rule.

nproellochs/ReinforcementLearning documentation built on March 3, 2020, 12:22 a.m.