Multinomial Naive Bayes, Python 3

Multinomial Naive Bayes

import collections as cl
import numpy as np

class MultinomialNB(object):

def __init__(self):
        self.X = None
        self.y = None

def __loading(self):
        self.list_labels = cl.Counter(self.Y)

for label in self.list_labels:
            temp = sum(self.X[self.Y==label]) + 1
            self.list_labels[label] = [
                    self.list_labels[label] / len(self.Y),
                    temp / sum(temp)
            ]

def fit(self, X, y):
        self.X = X
        self.Y = y
        self.__loading()

def predict(self, X):
        for x in X:
            _sum = 0
            for i in self.list_labels:
                temp = self.list_labels[i]
                self.list_labels[i] = (
                        temp[0] * np.prod(temp[1] ** x)
                )
                _sum += self.list_labels[i]
    
            yield {
                    item: self.list_labels[item] / _sum
                    for item in self.list_labels
            }

d1 = [2, 1, 1, 0, 0, 0, 0, 0, 0]
d2 = [1, 1, 0, 1, 1, 0, 0, 0, 0]
d3 = [0, 1, 0, 0, 1, 1, 0, 0, 0]
d4 = [0, 1, 0, 0, 0, 0, 1, 1, 1]

train_data = np.array([d1, d2, d3, d4])
label = np.array(['B', 'B', 'B', 'N'])

mnb = MultinomialNB()
mnb.fit(train_data, label)
print(list(mnb.predict([np.array([0, 1, 0, 0, 0, 0, 0, 1, 1])])))


run \| edit \| history \| help	0

λ

                                      .NET NoSQL database for rapid development