def train(self, spam_dir, ham_dir):
self.model = NTF_Model()
self.model.set_features(self.features)
N = 0
loss = 0.
for f in get_files(spam_dir):
print N
N += 1
self.model.observe_example(self.munge(f), 1)
for f in get_files(ham_dir):
print N
N += 1
self.model.observe_example(self.munge(f), 0)
self.model.build_network()
print "finished training"
class NB_NTF(NaiveBayesModel):
def classify(self, example, cost_ratio):
log_likelihood1 = math.log(self.model.base_param)
log_likelihood2 = math.log(1 - self.model.base_param)
for i, token in enumerate(self.model.attribute_params):
log_likelihood1 += (math.log(self.model.attribute_params[i][0]) -
example[i] / self.model.attribute_params[i][0])
log_likelihood2 += (math.log(self.model.attribute_params[i][1]) -
example[i] / self.model.attribute_params[i][1])
return int(log_likelihood1 - math.log(cost_ratio) > log_likelihood2)
def train(self, spam_dir, ham_dir):
self.model = NTF_Model()
self.model.set_features(self.features)
N = 0
loss = 0.
for f in get_files(spam_dir):
print N
N += 1
self.model.observe_example(self.munge(f), 1)
for f in get_files(ham_dir):
print N
N += 1
self.model.observe_example(self.munge(f), 0)
self.model.build_network()
print "finished training"
def munge(self, email_file):
f = open(email_file, 'rb')
text = f.read()
word_list = re.split('\W+', text)
num_words = len(word_list)
ntf_vector = [float(word_list.count(token)) /
num_words for token in self.features]
return ntf_vector
