Q-Learning算法 Matlab代码实现

state_num=6;
episode_num=300;

%初始化
gammer=0.8;  %给定折扣因子
reward_R=[
    -1-1-1-10-1;   %设置奖励矩阵R
    -1-1-10-1100;
    -1-1-10-1-1;
    -100-10-1;
    0-1-10-1100;
    -10-1-10100;
    ]
Q_table=zeros（size（reward_R））;  %设置初始化矩阵Q
final_state=6;
%For each episode
for i=1:episode_num
    current_state=randperm（state_num1）;   %随机选择一个状态
    while current_state~=final_state
    action_option=find（reward_R（