LDA的維度限制 - Githubissues

daengky / facerecog

Automatically exported from code.google.com/p/facerecog

0 stars 0 forks source link

Open GoogleCodeExporter opened 8 years ago

GoogleCodeExporter commented 8 years ago

為何在做LDA時，資料維度必須要大於資料量?

麻煩你解答了，謝謝~

Original issue reported on code.google.com by fishapp...@gmail.com on 1 Jun 2013 at 1:42

GoogleCodeExporter commented 8 years ago

我有段时间没碰这个项目了。不过我印象中LDA的样本数量是��
�以大于特征维度的，而且样本越多训练越准确啊

Original comment by xje...@gmail.com on 11 Oct 2013 at 6:56