search

uchicago-computation-workshop / marc_berman

0 stars 4 forks source link

Approach of calculating H and sources of cognitive effort #11

Open HaowenShang opened 6 years ago

HaowenShang commented 6 years ago

I'm interested in the approach to calculate the Hurst exponent (H). PSD and DFA are two approaches to calculate H. In principle, PSD may be used to calculate H. However, DFA values show a high degree of fit with large R^2 in this paper. I want to know whether DFA always has advantage comparing to PSD or it just obtains better model fit in this case? Also, cognitive effort is measured by task difficulty, task novelty, and aging effects. I wonder why other sources of cognitive effort are not taken into consideration? Are they all suppress fractal scaling of brain activity?

bermanm commented 6 years ago

DFA and PSD will typically yield similar results. PSD will have problems if you have a non-stationary signal and can also have problems at low-frequencies, which you can see from the plot of figure 2. Yes, there are many other sources of cognitive effort that were not explored, but we would hypothesize that they would all suppress fractal scaling.