Naive Bayes: Division by zero fix

mlfromscratch/examples/naive_bayes.py

@@ -5,11 +5,11 @@
 from mlfromscratch.supervised_learning import NaiveBayes
 
 def main():
-    data = datasets.load_iris()
+    data = datasets.load_digits()
     X = normalize(data.data)
     y = data.target
 
-    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.5)
+    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.4)
 
     clf = NaiveBayes()
     clf.fit(X_train, y_train)

mlfromscratch/supervised_learning/k_nearest_neighbors.py

@@ -2,7 +2,6 @@
 import numpy as np
 from mlfromscratch.utils import euclidean_distance
 
-
 class KNN():
     """ K Nearest Neighbors classifier.
 
@@ -40,3 +39,4 @@ def predict(self, X_test, X_train, y_train):
             label = self._vote(k_nearest_neighbors)
             y_pred[i] = label
         return y_pred
+

mlfromscratch/supervised_learning/naive_bayes.py

@@ -7,8 +7,6 @@
 
 class NaiveBayes():
     """The Gaussian Naive Bayes classifier. """
-    def __init__(self): pass
-
     def fit(self, X, y):
         self.X, self.y = X, y
         self.classes = np.unique(y)
@@ -26,8 +24,9 @@ def fit(self, X, y):
 
     def _calculate_likelihood(self, mean, var, x):
         """ Gaussian likelihood of the data x given mean and var """
-        coeff = (1.0 / (math.sqrt((2.0 * math.pi) * var)))
-        exponent = math.exp(-(math.pow(x - mean, 2) / (2 * var)))
+        eps = 1e-4 # Add small term in denominator to avoid division by zero
+        coeff = 1.0 / (math.sqrt((2.0 * math.pi) * var) + eps)
+        exponent = math.exp(-(math.pow(x - mean, 2) / (2 * var  + eps)))
         return coeff * exponent
 
     def _calculate_prior(self, c):
@@ -77,4 +76,3 @@ def predict(self, X):
             y = self._classify(sample)
             y_pred.append(y)
         return y_pred
-