User:Freeman

From Wikiversity

Jump to: navigation, search

[edit] Who?

My name is Michael Joya. I'm a 3rd year computer science major specializing in cognitive systems at the University of British Columbia. My current interests are in data mining and machine learning. I'm currently looking at data mining and visualization methods for neuroimaging data such as that generated by fMRI and EEG/ERP recordings.

[edit] What?

Today it is rather common for neuroscientists to publish aggregate data of the brain activity that is activated by a particular task. To obtain consensus on localized regions, researchers must use a combination of averaging and subtraction techniques on the data from many subjects. These cancel out noise and form contrasts between groups to produce an aggregate of the activity. The output of these techniques delivers a summary picture that represents the entire population.

It cannot be known whether these aggregates are good representations of the overall subject pool without some consideration for the variance in the population, but this is not accounted for by current methodologies. It is my current task, therefore, to implement a visualization method that represents the means, variances, and intersections in all subgroups, however disparate these may be.

[edit] Links