Inferring Trajectories of Psychotic Disorders Using Dynamic Causal Modeling.

Finally, leave-one-out cross-validation using the 24 virtual subjects from groups 1 and 5 ( Figure 4C ) showed a highly significant out-of-samples correlation between the estimated group membership and the actual group membership, demonstrating that the effect size was large enough for the Rayleigh parameter to have clinical relevance (in this simulation).

also in 132,142 other papers