Fig 1.
Visualisation of the DPNN architecture.
Table 1.
High variance across prototypes.
Table 2.
Low variance within prototypes.
Fig 2.
Visualisation of L3’s computation.
Table 3.
Experiment 1 (synthetic data): Median scores for each evaluated method (column) and evaluation metric (row).
Table 4.
Experiment 2 (clinical data): Median scores for each evaluated method (column) and evaluation metric (row).
Table 5.
Hyper-parameters of the DPNN model in both experiments.
Table 6.
Train-test ratio influence on the performance of the DPNN model with the MDD data, medians results.
Table 7.
Sensitivity analysis- synthetic data.
Performance by Number of Clusters.
Table 8.
Sensitivity analysis- MDD data.
Performance by Number of Clusters.