art_complement_code.m

来自「ART 神经网络Matlab代码。包括数据产生函数」· M 代码 · 共 40 行

function complementCodedData = ART_Complement_Code(data)
% ART_Complement_Code    Complement codes the data for use with an ART network.
%    COMPLEMENTCODEDDATA = ART_Complement_Code(DATA)
%    This function complement codes the given data where the complement
%    of x is 1-x. For example, suppose the data were as follows:
%           data = 0.3  0.2  0.6
%                  0.4  0.1  0.8
%    Then the complement coding would be as follows:
%           complementCodedData = 0.3  0.2  0.6
%                                 0.7  0.8  0.4
%                                 0.4  0.1  0.8
%                                 0.6  0.9  0.2
% 
%    The input parameters are as follows:
%    The DATA is a matrix of size NumFeatures-by-NumSamples that holds
%    the data to be complement coded. The complement coding works on
%    each column (i.e., each sample).
%
%    The return parameter is as follows:
%    The COMPLEMENTCODEDDATA is the data that has been complement coded.
%    It is a matrix of size 2*NumFeatures-by-NumSamples.


% Determine the size of the data.
[numFeatures, numSamples] = size(data);

% Create the return variable.
complementCodedData = ones(2*numFeatures, numSamples);

% Do the complement coding for each sample.
for j = 1:numSamples
    count = 1;
    for i = 1:2:(2*numFeatures)
        complementCodedData(i, j) = data(count, j);
        complementCodedData(i + 1, j) = 1 - data(count, j);
        count = count + 1;
    end
end

return