Hyperparameter optimization for GPCA predictions
Philipp Trepte
2023-10-25
Hyperparameter comparison (e, i, C) for the maSVM algorithm trained on the reference set
Hyperarameters
ensembleSize: 25
iter: 1
C: 0.01
(A) Recovery training data, (B) ROC curve training data
C: 0.1
(A) Recovery training data, (B) ROC curve training data
iter: 5
C: 0.01
(A) Recovery training data, (B) ROC curve training data
C: 0.1
(A) Recovery training data, (B) ROC curve training data
iter: 10
C: 0.01
(A) Recovery training data, (B) ROC curve training data
C: 0.1
(A) Recovery training data, (B) ROC curve training data
ensembleSize: 50
iter: 1
C: 0.01
(A) Recovery training data, (B) ROC curve training data
C: 0.1
(A) Recovery training data, (B) ROC curve training data
iter: 5
C: 0.01
(A) Recovery training data, (B) ROC curve training data
C: 0.1
(A) Recovery training data, (B) ROC curve training data
iter: 10
C: 0.01
(A) Recovery training data, (B) ROC curve training data
C: 0.1
(A) Recovery training data, (B) ROC curve training data
ensembleSize: 100
iter: 1
C: 0.01
(A) Recovery training data, (B) ROC curve training data
C: 0.1
(A) Recovery training data, (B) ROC curve training data
iter: 5
C: 0.01
(A) Recovery training data, (B) ROC curve training data
C: 0.1
(A) Recovery training data, (B) ROC curve training data
iter: 10
C: 0.01
(A) Recovery training data, (B) ROC curve training data
C: 0.1
(A) Recovery training data, (B) ROC curve training data