The Spider Objects

 
 
 
 
   
  Random Forest Decision Tree  by L.Breiman  [WEKA Required]
   
 


   Original WEKA Code from Richard Kirkby
   Please see WEKA: http://www.cs.waikato.ac.nz/ml/weka/
  
   Generates a random forest wrapper on the WEKA Random Forest implementation with given hyperparameters.
  
  
     Hyperparameters (with defaults)
     numoftrees=10      -- set to number of trees
     numoffeatures=0  -- number of features considered. 0 means use all, -1
     use only log Nmax features (Nmax equals input dimension)
   
     seed=0           -- used seed
    Model
      tree               -- the tree
  
   Methods:
    train, test, plot 
    
   Example:
  d=gen(toy2d('2circles','l=200'));
  r0=randomforest;
  [cr,ca]=train(cv(r0),d); loss(cr), pause
  [r,a]=train(s0,d);
  disp('probability output'), plot(a,d), pause,