Skip to main content

Advertisement

Table 7 Multiple Linear Regression Model Predicting SWMT Performance by SVM Model 1 Features

From: Machine learning identification of EEG features predicting working memory performance in schizophrenia and healthy adults

Step R R2 Std. Error Δ R2 Δ F Δ F p value
1 .450a .202 9.740 .202 12.661 .001
2 .655b .428 8.327 .226 19.410 .000
3 .832c .692 6.177 .263 41.051 .000
4 .873d .762 5.484 .070 13.894 .001
  1. Multiple linear regression based on all SVM Model 1 features achieved maximum fit based on frontal gamma during encoding and central theta 1 during retention. Features selected using forward-stepwise entry in the following order: 1 Central Retain Theta 1 – correct; 2 (1) + Frontal Encode Gamma – correct; 3 (1, 2) + Frontal Encode Gamma – incorrect, 4 Predictors (1, 2, 3) + Central Retain Theta 1 – incorrect