Gaussian Naive Bayes for text: Spam Filter for e-mail

ナイーブベイズ分類器によるEメールスパムフィルター

Introduction

The equation at the center of naive Bayes is an application of Bayes theorem.

The posterior probability P (cat | doc) is the probability that it is the category cat given the document doc. Unknown documents for which you want to predict categories are classified into categories with the highest posterior probability (MAP estimation). In order to calculate this probability, the prior probability P (cat) on the right side and the likelihood P (doc | cat) are required. P (doc) is common to all categories and can be ignored. The posterior probability P (cat | doc) and the likelihood P (doc | cat) are confusing but different.

Technical Preferences

Title	Detail
Environment	MacOS Mojave 10.14.3
Language	Python
Library	scikit-learn, Numpy, matplotlib, Pandas, Seaborn
Dataset	Twitter
Algorithm	Gaussian Naive Bayes

Refference

• Understanding Naive Bayes and its application in text classification
• 6 Easy Steps to Learn Naive Bayes Algorithm
• Implementing a Naive Bayes classifier for text categorization in five steps
• Text Classification in NLP — Naive Bayes