reinforcement learning example in matlab

main.m
% Define the environment
rewardMatrix = [0 0 0 0 0 0 0 0 0 100;
                0 0 0 0 0 0 0 0 0 0;
                0 0 0 0 0 0 0 0 0 0;
                0 0 0 0 0 0 0 0 0 0;
                0 0 0 0 0 0 0 0 0 0;
                0 0 0 0 0 0 0 0 0 0;
                0 0 0 0 0 0 20 0 0 0;
                0 0 0 0 0 0 0 0 0 0;
                0 0 0 0 0 0 0 0 0 0;
                0 0 0 0 0 0 0 0 0 10];

numStates = size(rewardMatrix, 1);
numActions = size(rewardMatrix, 2);

% Initialize the Q-table
Q = zeros(numStates, numActions);

% Set the hyperparameters
gamma = 0.8;  % discount factor
alpha = 0.2;  % learning rate
numEpisodes = 1000;

% Run the Q-learning algorithm
for episode = 1:numEpisodes
    % Initialize the starting state
    currentState = randi(numStates);
    
    while currentState ~= numStates
        % Choose an action using epsilon-greedy policy
        epsilon = 0.2;  % exploration rate
        if rand() < epsilon
            action = randi(numActions);
        else
            [~, action] = max(Q(currentState, :));
        end
        
        % Update the Q-table
        nextState = action;
        reward = rewardMatrix(currentState, action);
        Q(currentState, action) = Q(currentState, action) + alpha * (reward + gamma * max(Q(nextState, :)) - Q(currentState, action));
        
        % Transition to the next state
        currentState = nextState;
    end
end

% Use the learned Q-table to choose the optimal policy
optimalPolicy = zeros(numStates, 1);
for state = 1:numStates
    [~, optimalPolicy(state)] = max(Q(state, :));
end
1576 chars
53 lines