On the improvement of schizophrenia detection with optical coherence tomography data using deep neural networks and aggregation functions

Figures

• 

  Figure 1. General overall scheme of the aggregation procedure.

• 

  Figure 2. Accuracy of the Quadrature-Inspired Smooth Generalized Choquet Integral in dependence on the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha$$\end{document} parameter value.

• 

  Figure 3. Accuracy of the classical pre-aggregation function in dependence on the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha$$\end{document} parameter value.

• 

  Figure 4. Standard deviations in dependence on the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha$$\end{document} parameter values.

• 

  Figure 5. ROC curves with the AUC values for the aggregation operators denoted by the formulae numbers.