shared_page_1

The visualization of the biclustering results from Sonya:

Experiment#1: Comparing the biclustering algorithms and the number of bicluster k

Each voxel is described by the descriptor "Hypothesis-OAO".

The biclustering algorithm is run on both hemispheres simultaneously.

Progress: The data is already done, Sonya is working on the visualization.

We want to find out:

1) Which algorithms give better (homogeneous region in the brain) result?

2) Which is a good k

Experiment#2: Biclustering on the descriptors or on the beta, which is better?

Hypo-OAO --> use the results from experiment#1

Beta coefficients -- Bot is working on this result

Experiment#3: The difference of biclustering on 1) left, 2) right and 3) left-right simultaneously

just pick a good k={4,6,8}, use {CoCluster or nmfrule}