The Spider Objects

 
 
 
 
   
  L1 norm for linear classifiers.
   
 


   minimize l1 norm , i.e |w|_1 
      under separability contraints (w.x_i+b)>1
    
   A=L1(H) returns an l1 object initialized with hyperparameters H. 
  
   Hyperparameters, and their defaults:
   
   Model:
    alpha                -- the weights of the hyperplane
    b0                   -- the threshold
  
   Methods:
    train, test