How to check independence assumption for Naive Bayes?

2 次查看（过去 30 天）
Isra Abuhasna
Isra Abuhasna 2018-9-22
评论： Walter Roberson 2018-9-25
I know that the assumption for the Naive Bayes classifier is that the features are independent, but what if we want to check if that applies to our data? How would we do test it using Matlab?

请先登录，再回答此问题。

采纳的回答

Zenin Easa Panthakkalakath
For a 2-dimensional matrix, if the determinant is zero, then at least two rows or two columns are linearly dependent.
You can also do a principal component analysis in order to know in which direction the variation is significant.
  1 个评论
Walter Roberson
Walter Roberson 2018-9-25
The determinant would be 0 for non-square matrices.
You could check whether rank() was min() rows, cols, with anything less implying dependence beyond the usual.

请先登录，再进行评论。

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Naive Bayes 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by