这部分主要是参考教程后写的代码，比较简单易懂

    1. clear
    2. clc
    3. close all
    4. if strcmpi(computer, 'PCWIN') | strcmpi(computer, 'PCWIN64')
    5.     compile_windows
    6. else
    7.     compile_linux
    8. end
    10. totalTrainTime = 0;
    11. totalTestTime = 0;
    12. %---------------------载入数据--------------------
    13. bedroom = load('C:\Users\*****\Desktop\数据集\bedroom.mat');
    14. bedroomIndex = fieldnames(bedroom);
    15. bedroom = bedroom.(bedroomIndex{1});
    16. forest = load('C:\Users\*****\Desktop\数据集\forest.mat');
    17. forestIndex = fieldnames(forest);
    18. forest = forest.(forestIndex{1});
    19. label = load('C:\Users\*****\Desktop\数据集\labelset.mat');
    20. labelIndex = fieldnames(label);
    21. label = label.(labelIndex{1});
    23. %-------------将数据混合并且打乱-------------
    24. xData = [bedroom; forest];
    25. [N, D] = size(xData);
    26. randVector = randperm(N);
    27. xTrn = xData(randVector(1:15), :);  
    28. yTrn = label(randVector(1:15));
    29. xTst = xData(randVector(16:end), :);
    30. yTst = label(randVector(16:end));
    32. %------------训练数据集---------------
    33. model = classRF_train(xTrn, yTrn, 500, 2);
    34. yHat = classRF_predict(xTst, model);
    35. fprintf('\n 错误率%f\n', length(find(yHat ~= yTst))/length(yTst));
    37. %------------绘出训练错误曲线------------
    38. figure('Name','OOB error rate');
    39. plot(model.errtr(:,1)); title('OOB error rate');  
    40. xlabel('iteration (# trees)'); ylabel('OOB error rate');
    41. %{因为随机森林属于bagging类的集成学习器，所以最后结果是投票决定的，
    42. 这个错误曲线是每个基学习器的分类错误曲线%}

    数据集.zip