Table 5 Confusion matrix for the SVM algorithm for different rs-fMRI metrics (ADNI dataset): confusion matrices that show the performance of the GA, SFFS, PSO algorithms when used with the SVM classifier.

From: Comparison of the diagnostic accuracy of resting-state fMRI driven machine learning algorithms in the detection of mild cognitive impairment

GA-LCOR

SFFS-LCOR

PSO-LCOR

SFFS-concatenated metrics

Confusion matrix

Sensitivity

Specificity

Confusion matrix

Sensitivity

Specificity

Confusion matrix

Sensitivity

Specificity

Confusion matrix

Sensitivity

Specificity

\(\left[\begin{array}{cc}80& 6\\ 10& 59\end{array}\right]\)

93%

85%

\(\left[\begin{array}{cc}78& 8\\ 8& 61\end{array}\right]\)

90%

88%

\(\left[\begin{array}{cc}79& 7\\ 9& 60\end{array}\right]\)

91%

86%

\(\left[\begin{array}{cc}74& 12\\ 4& 65\end{array}\right]\)

89%

94%

  1. The table also shows the sensitivity and specificity values for each depicted algorithm.
  2. GA genetic algorithm, SFFS sequential floating forward selection, NSGA-II non-dominated sorting genetic algorithm II, PSO particle swarm optimization, SVM support vector machine.
Back to article page