How to use PCA and k-means for clustering excel file datasest in Matlab?

3 views (last 30 days)
I have an excel file containing normalized EEG dataset features for some disease (Healthy and unhealthy patient details).
It has a 301x16 dimension where 1st row and 1st column contain feature names and dataset names respectively; that means numerical values are within 300x15 dimension (300 dataset with 15 unique features)
I want to classify these datasets into clusters as: 1st 100 rows must go in one cluster, next 100 in 2nd cluster and last 100 rows in 3rd cluster (as 1st hundreds contain features of healthy patient data, 2nd hundred one type of disease and last hundred 2nd type of disease).
Please help me regarding whether it is possible to do this task with PCA and if Yes then how it can be used for this purpose in Matlab (I have Matlab R2010a).
and how further K-means clustering can be applied to show clusters. later I need to check accuracy.
  1 Comment
Image Analyst
Image Analyst on 30 Apr 2017
OK, how can we help, given that you have not given us any data to work with? About all I can say is to follow the examples in the help for pca() and kmeans(). If you want more, then post your data along with some code to get it into a variable in MATLAB.

Sign in to comment.

Answers (0)

Community Treasure Hunt

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

Start Hunting!