Table 6 Different results while using different parameters.

From: Explainable AI-based suicidal and non-suicidal ideations detection from social media text with enhanced ensemble technique

Parameters

Metrics

 

SVM

LR

GB

DT

Stacking

Average Score

SVM:

C = 50

Gamma = auto

Probability = true

Degree = 1

LR:

C = 200

GB:

Max_features = 0.2

Max_depth = 10

Min_samples_leaf = 2

DT:

Max_depth = 6

random_state = 42

StackingClassifier

cv = 3, estimators = [('SVM,LR,GB,DT

final_estimator = RF(Max_features = auto

Max_depth = 10

Min_samples_leaf = 0.005

Min_samples_split = 0.005

n_jobs = -1

n_estimators = 10

random_state = 40

Accuracy

Suicidal

0.52

0.90

0.91

0.73

0.94

92%

Non-Suicidal

0.70

0.94

0.95

0.81

0.94

Precision

Suicidal

0.92

0.87

0.91

0.76

0.95

91%

Non-Suicidal

0.97

0.98

0.98

0.79

0.98

Recall

Suicidal

0.73

0.94

0.92

0.75

0.96

94%

Non-Suicidal

0.48

0.93

0.93

0.83

0.97

F1-Score

Suicidal

0.67

0.90

0.91

0.72

0.94

91%

Non-Suicidal

0.62

0.95

0.95

0.81

0.91