Source Code for Module pycv.cs.ml.cla.nb

 1  # PyCV - A Computer Vision Package for Python Incorporating Fast Training of Face Detection 
 2   
 3  # Copyright 2007 Nanyang Technological University, Singapore. 
 4  # Authors: Minh-Tri Pham, Viet-Dung D. Hoang, and Tat-Jen Cham. 
 5   
 6  # This file is part of PyCV. 
 7   
 8  # PyCV is free software: you can redistribute it and/or modify 
 9  # it under the terms of the GNU General Public  
10  # License as published by the Free Software Foundation, either version  
11  # 3 of the License, or (at your option) any later version. 
12   
13  # PyCV is distributed in the hope that it will be useful, 
14  # but WITHOUT ANY WARRANTY; without even the implied warranty of 
15  # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the 
16  # GNU General Public License for more details. 
17   
18  # You should have received a copy of the GNU General Public License 
19  # along with this program.  If not, see <http://www.gnu.org/licenses/>. 
20   
21  # --------------------------------------------------------------------- 
22  #!/usr/bin/env python 
23   
24   
25  __all__ = ['NBClassifier','train_NBClassifier'] 
26   
27  from copy import copy 
28  from math import fabs, log, exp 
29  from numpy import array, where, zeros, ones, dot, prod 
30  from numpy import log as NP_log 
31  from numpy import exp as NP_exp 
32  from numpy import sqrt as NP_sqrt 
33   
34  from cla import Classifier 
35  from dataset import WeightedCDataset 
36   
37  from pycv.cs.ml import PredictPdfInterface, OnlineLearningInterface 
38  from pycv import tprint 
39  from pycv.ext import nb_NBClassifier_predict_pdf 
40   
41  #------------------------------------------------------------------------------- 
42  # Naive Bayes Classifier 
43  #------------------------------------------------------------------------------- 
44   
45 -class NBClassifier(PredictPdfInterface, OnlineLearningInterface, \ 
46      Classifier): # Naive Bayes classifier 
47   
48 -    def __init__(self,nclasses,stats2e): 
49          Classifier.__init__(self,nclasses) 
50          self.stats2e = stats2e 
51   
52 -    def predict_pdf(self, input_point, *args, **kwds): 
53          ip = input_point 
54          pdf = zeros(self.nclasses) 
55   
56          #gstat = self.stats2e.get_cond_stat 
57          # for j in xrange(self.nclasses): 
58              # pdf[j] = NP_log(gstat(j,0).squeeze()) 
59              # pdf[j] -= 0.5 * (((ip - gstat(j,1)) / gstat(j,2))**2).ravel().sum() 
60              # pdf[j] -= NP_log(gstat(j,2)).sum() 
61          # pdf = NP_exp(pdf) 
62          #return pdf/pdf.sum() 
63           
64          A = self.stats2e.A 
65          d = len(ip) 
66          J = self.nclasses 
67          nb_NBClassifier_predict_pdf(ip, pdf, A, d, J) 
68          return pdf 
69           
70 -    def learn(self, input_point, j, w = None, *args, **kwds): 
71          """Learn incrementally with a new input point, its class, and optionally its weight. 
72   
73          Input: 
74              input_point: a new input point 
75              j: its corresponding class 
76              w: optionally its weight, or 1 if not specified 
77          """ 
78          self.stats2e.learn(input_point, j, w) 
79   
80   
81 -def train_NBClassifier( cd ): 
82      """Train a NBClassifier using a WeightedCDataset 
83   
84      Input: 
85          classification_dataset: a WeightedCDataset 
86      Output: 
87          an NBClassifier 
88      """ 
89      return NBClassifier(cd.J,cd.compute_Stats2e()) 
90