Discrimination between samples
0
0
Entering edit mode
7.9 years ago
forever ▴ 80

Hi everyone, I have 100 samples x100 features dataset. The 100 samples are belong to 4 different class. PCA can not show any difference between these 4 classes. Also, k-means only shows 2 clusters. What do you suggest me to do further? I appreciate your help.

R • 1.3k views
ADD COMMENT
0
Entering edit mode

Perhaps you could explain in more depth? As it stands, your question is unrelated to bioinformatics.

ADD REPLY
0
Entering edit mode

I used k-means and NMF and they show only 2 clusters. I want to analysis my dataset as it produced from different sources. I expect to see 4 different patterns, maybe you can suggest me something else other than PCA, k-means and NMF?

ADD REPLY
0
Entering edit mode

tSNE? But really, the question of what you're actually trying to do still stands. If you know already that the samples come from 4 different sources then you don't need to find that out with PCA/k-means/etc. At most, you're just finding out that the sources produces largely indistinguishable samples as assayed by your 100 features.

ADD REPLY
0
Entering edit mode

Just checking: Did you set k to 4?

ADD REPLY
0
Entering edit mode

Hi forever,

first please avoid 'blinded' questions, as such do not serve a purpose for neither you nor the community. Your problem description is insufficient to provide a good answer. Please state the nature of your data,

  • experiment type
  • data type and origin
  • pre-processing/transformation

    If these are not of biological origin, your question does not fit here.

ADD REPLY

Login before adding your answer.

Traffic: 1750 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6