Table 4 Prediction results for individual classifier algorithms.

From: An open automation system for predatory journal detection

Model

Number of Word Features (NWF)

Recall

Number of Word Features (NWF)

F1

GNB

8450

0.89

3700

0.752

MNB

1000

0.904

1150

0.93

Logistic Regression

350

0.964

1650

0.97

Random Forest

850

0.982

1200

0.98

SGD

7950

0.97

1550

0.972

SVM

350

0.952

2400

0.934

KNN

3000

0.945

500

0.931

Voting

2900

0.97

1700

0.973

Voting (no GNB)

2150

0.976

1100

0.972

Voting (the top 3 model)

950

0.94

1800

0.975

  1. GNB, Gaussian naïve Bayes; MNB, multinomial naïve Bayes; SGD, stochastic gradient descent; SVM, support vector machine; KNN, K-nearest neighbor.
Back to article page