random_dd.m

模式识别工具箱,希望对大家有用!

%RANDOM_DD Random one-class classifier
% 
%       W = random_dd(A,fracrej)
% 
% This is the trivial one-class classifier, randomly assigning labels
% and rejecting fracrej of the data objects. This procedure is just to
% show the basic setup of a prtools classifier, and what is required
% to define a one-class classifier.

% Copyright: D. Tax, R.P.W. Duin, [e-mail]
% Faculty of Applied Physics, Delft University of Technology
% P.O. Box 5046, 2600 GA Delft, The Netherlands
  
function W = empty_dd(a,fracrej)

if nargin < 2 fracrej = 0.05; end
if nargin < 1 | isempty(a) 
  W = mapping(mfilename,{fracrej});
  return
end

if isa(fracrej,'double')           %training

  if ~isa(a,'dataset')              %train on training set
    error('I need a dataset to train');
  end
  a = target_class(a);     % only use the target class
  [nlab,lablist,m,k,c,prob,featlist] = dataset(a);

  % train it:

  % this trivial classifier cannot be trained. for each object we will
  % output a random value between 0 and 1, indicating the probability
  % that an object belongs to class 'target'
  % if we would like to train something, we should do it here.

  %and save all useful data:
  W.threshold = fracrej;  % a threshold should always be defined
  W = mapping(mfilename,W,str2mat('target','outlier'),0,c);

else                               %testing

  [nlab,lablist,m,k,c,p,featlist] = dataset(a);
  [W,classlist,type,k,c] = mapping(fracrej);  % unpack

  % output should consist of two numbers: the first indicating the
  % probability that it belongs to the target, the second indicating
  % the probability that it belongs to the outlier class. The latter
  % is often the constant threshold:
  newout = [rand(m,1) ones(m,1)*W.threshold];

  W = dataset(newout,getlab(a),classlist,p,lablist);
end
return