matlab中cart树,matlab cart决策树

function D = CART(train_features, train_targets, params, region)

% Classify using classification and regression trees

% Inputs:

%         features        - Train features

%        targets            - Train targets

%        params                - [Impurity type, Percentage of incorrectly assigned samples at a node]

%                   Impurity can be: Entropy, Variance (or Gini), or Missclassification

%        region            - Decision region vector: [-x x -y y number_of_points]

%

% Outputs

%        D                        - Decision sufrace

[Ni, M]                   = size(train_features);

%Get parameters

[split_type, inc_node] = process_params(params);

%For the decision region

N           = region(5);

mx          = ones(N,1) * linspace (region(1),region(2),N); %linspace(x1,x2,n)  用于生成x1,x2之间个数为n的向量

my          = linspace (region(3),region(4),N)' * ones(1,N); %ones(m,m) 用于生成全部为1的m*n矩阵

flatxy      = [mx(:), my(:)]';

%Preprocessing

[f, t, UW, m]   = PCA(train_features, train_targets, Ni, region);

train_features  = UW * (train_features - m*ones(1,M));

flatxy          = UW * (flatxy - m*ones(1,N^2));

%Build the tree recursively

disp('Building tree')

tree        = make_tree(train_features, train_targets, M, split_type, inc_node, region);

%Make the decision region according to the tree

disp('Building decision surface using the tree')

targets                = use_tree(flatxy, 1:N^2, tree);

D                                = reshape(targets,N,N);

%END

function targets = use_tree(features, indices, tree)

%Classify recursively using a tree

if isnumeric(tree.Raction)

%Reached an end node

targets = zeros(1,size(features,2));

targets(indices) = tree.Raction(1);

else

%Reached a branching, so:

%Find who goes where

in_right    = indices(find(eval(tree.Raction)));

in_left     = indices(find(eval(tree.Laction)));

Ltargets         = use_tree(features, in_left, tree.left);

Rtargets         = use_tree(features, in_right, tree.right);

targets                 = Ltargets + Rtargets;

end

%END use_tree

function tree = make_tree(features, targets, Dlength, split_type, inc_node, region)

%Build a tree recursively

if (length(unique(targets)) == 1),

%There is only one type of targets, and this generates a warning, so deal with it separately

tree.right      = [];

tree.left       = [];

tree.Raction    = targets(1);

tree.Laction    = targets(1);

else

end

[Ni, M] = size(features);

Nt      = unique(targets);

N       = hist(targets, Nt);

if ((sum(N < Dlength*inc_node) == length(Nt) - 1) || (M == 1)),

%No further splitting is neccessary

tree.right      = [];

tree.left       = [];

if (length(Nt) ~= 1),

MLlabel                          = find(N == max(N));

else

MLlabel                   = 1;

end

tree.Raction    = Nt(MLlabel);

tree.Laction    = Nt(MLlabel);

else

%Split the node according to the splitting criterion

deltaI                = zeros(1,Ni);

split_point = zeros(1,Ni);

op                                = optimset('Display', 'off');

for i = 1:Ni,

split_point(i) = fminbnd('CARTfunctions', region(i*2-1), region(i*2), op, features, targets, i, split_type);

I(i)                                = feval('CARTfunctions', split_point(i), features, targets, i, split_type);

end

[m, dim]                                = min(I);

loc                                        = split_point(dim);

%So, the split is to be on dimention 'dim' at location 'loc'

indices                 = 1:M;

tree.Raction= ['features(' num2str(dim) ',indices) >  ' num2str(loc)];

tree.Laction= ['features(' num2str(dim) ',indices) <= ' num2str(loc)];

in_right    = find(eval(tree.Raction));

in_left     = find(eval(tree.Laction));

if isempty(in_right) || isempty(in_left)

%No possible split found

tree.right      = [];

tree.left       = [];

if (length(Nt) ~= 1),

MLlabel           = find(N == max(N));

else

MLlabel                  = 1;

end

tree.Raction    = Nt(MLlabel);

tree.Laction    = Nt(MLlabel);

else

%...It's possible to build new nodes

tree.right = make_tree(features(:,in_right), targets(in_right), Dlength, split_type, inc_node, region);

tree.left  = make_tree(features(:,in_left), targets(in_left), Dlength, split_type, inc_node, region);

end

end