search

srikanthguduru / imrmc

Automatically exported from code.google.com/p/imrmc
0 stars 0 forks source link

Variance estimation in data with more than one truth file. #28

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
I am using a set of 600 trial images for each experiment, which can be divided 
in two groups. The first one of 305 trial images that test higher levels of a 
color scale and a second group of 295 trial images that test the lower levels. 
This two groups are different images and have different truth. I would be 
interested in calculating the variance of the performance of a combination of 
color/device between both groups. Any pointers? Thank you.

Original issue reported on code.google.com by silv...@gmail.com on 4 Mar 2015 at 12:32

GoogleCodeExporter commented 9 years ago 
You can label them as coming from different modalities. The cases are different 
across modalities, but this is not a fundamental problem. It does increase your 
variance. There is no correlation from the cases in this situation; correlation 
helps when you are taking a difference across modalities.

Original comment by Brandon.Gallas on 7 Mar 2015 at 6:21

GoogleCodeExporter commented 9 years ago 
Removing priority so that answered questions go to the bottom of the issues 
list.

Original comment by Brandon.Gallas on 25 Mar 2015 at 5:47