Skip to main content

Table 4 Observed differences between the testing results and each language with p values from permutation tests

From: Comparison between machine learning methods for mortality prediction for sepsis patients with different social determinants

 

English

Other

Spanish

 

Observed difference

p_val

Observed difference

p_val

Observed difference

p_val

Ridge classifier

0.0154

0.279

− 0.0760

0.107

− 0.3422

0.012

Perceptron

0.0182

0.252

− 0.0916

0.053

− 0.3551

0.004

Passive-aggressive

0.0122

0.301

− 0.0555

0.172

− 0.2288

0.053

kNN

0.0166

0.263

− 0.0768

0.102

− 0.3063

0.017

Random forest

0.0037

0.409

− 0.0057

0.489

− 0.2342

0.002

LinearSVC_L1

0.0160

0.299

− 0.0772

0.102

− 0.3428

0.003

LinearSVC_L2

0.0156

0.297

− 0.0763

0.121

− 0.3424

0.007

SGDClassifier_L1

0.0184

0.246

− 0.0783

0.093

− 0.3347

0.008

SGDClassifier_L2

0.0187

0.269

− 0.0752

0.107

− 0.3396

0.004

SGDClassifier_EN

0.0181

0.259

− 0.0760

0.105

− 0.3283

0.006

MultinomialNB

0.0221

0.224

− 0.1210

0.021

− 0.2746

0.031

BernoulliNB

0.0076

0.389

− 0.0621

0.082

0.0306

0.422

Logistic regression

0.0145

0.293

− 0.0703

0.125

− 0.3173

0.014

SVC_rbf

0.0159

0.306

− 0.0825

0.080

− 0.3332

0.012

SVC_poly

0.0176

0.275

− 0.0860

0.079

− 0.3633

0.002

SVC_sigmoid

− 0.0030

0.454

0.0341

0.288

− 0.1814

0.089

  1. Observe difference: observed difference in AUC when compared with the performance on the entire testing set; p_val: p-value, p values less than or equal to 0.05 were highlighted; Passive-aggressive: passive-aggressive classifier; kNN: k-Nearest Neighbors; LinearSVC_L1 or _L2: support vector machine with linear kernel coupled with L1 or L2 regularization; SGDClassifier_L1 or _L2 or _EN: stochastic gradient descent with L1 or L2 or Elastic Net regularization; MultinomialNB: Multinomial naïve Bayes; BernoulliNB: Bernoulli naïve Bayes; SVC_rbf or _poly or _sigmoid: support vector machine with rbf kernel or polynomial kernel or sigmoid kernel