C# Класс numl.Reinforcement.QLearning.QLearnerModel

A Q-Learner Model.
Наследование: ReinforcementModel
Показать файл Открыть проект

Открытые методы

Метод Описание
Learn ( Vector x1, double y, Vector x2, double r ) : void

Updates the Q-Learner model by reinforcing with the new state/action and transition state feedback values.

Learn ( Vector x, double y, double r ) : void

Updates the Q-Learner model by reinforcing with the new state/action feedback values.

Predict ( Vector y ) : double

Predicts the best action for the current state.

QLearnerModel ( ) : System

Initializes a new Q-Learner model.

Описание методов

Learn() публичный Метод

Updates the Q-Learner model by reinforcing with the new state/action and transition state feedback values.
public Learn ( Vector x1, double y, Vector x2, double r ) : void
x1 numl.Math.LinearAlgebra.Vector Item features, i.e. the original State.
y double Action label.
x2 numl.Math.LinearAlgebra.Vector Transition state value.
r double Reward value.
Результат void

Learn() публичный Метод

Updates the Q-Learner model by reinforcing with the new state/action feedback values.
public Learn ( Vector x, double y, double r ) : void
x numl.Math.LinearAlgebra.Vector State vector.
y double Action label.
r double Reward value.
Результат void

Predict() публичный Метод

Predicts the best action for the current state.
public Predict ( Vector y ) : double
y numl.Math.LinearAlgebra.Vector
Результат double

QLearnerModel() публичный Метод

Initializes a new Q-Learner model.
public QLearnerModel ( ) : System
Результат System