A custom implementation of the Gaussian Naive Bayes classifier that I created myself to study how the algorithm works. Comes very close to matching the scikit learn implementation's predictions.
[[ 61 6]
[ 6 115]]
0.9361702127659575
my custom GNB: Number of mislabeled points out of a total 188 points : 12
[[ 61 6]
[ 5 116]]
0.9414893617021277
scikit learn GNB: Number of mislabeled points out of a total 188 points : 11