Additionally, the words in the message body are not independent from one another, since the appearance of some words is a very good indication that other words are also likely to appear.
- 4. Probabilistic Learning – Classification Using Naive Bayes
- from Machine Learning with R - Second Edition
- Publisher: Packt Publishing
- Released: July 2015
example implies not all features are independent
Share this highlighthttp://www.safaribooksonline.com/a/machine-learning-with/10232797/