Table 8 Performance assessment of using the neural networks (NN) models.

From: Machine learning-based prediction of drug response in ischemia reperfusion animal model

Model

Neural networks (NN)

Metric

Complete

(all features)

Reduced model

(SFS)

Molecular

Accuracy

0.8766 ∓ 0.0200

0.7702 ∓ 0.0827

 

Precision

0.9176 ∓ 0.0202

0.9273 ∓ 0.0280

 

Recall

0.9025 ∓ 0.0503

0.7250 ∓ 0.1230

 

Specificity

0.8222 ∓ 0.0480

0.8759 ∓ 0.0489

 

MCC

0.7228 ∓ 0.0311

0.5699 ∓ 0.1327

Biochemical

Accuracy

0.8541 ∓ 0.0274

0.9103 ∓ 0.0203

 

Precision

0.9325 ∓ 0.0478

0.9570 ∓ 0.0118

 

Recall

0.8538 ∓ 0.0504

0.9107 ∓ 0.0283

 

Specificity

0.8593 ∓ 0.0999

0.9111 ∓ 0.0240

 

MCC

0.6887 ∓ 0.0721

0.8009 ∓ 0.0455

Molecular-biochemical

Accuracy

0.8653 ∓ 0.0473

0.9103 ∓ 0.0203

 

Precision

0.9086 ∓ 0.0218

0.9570 ∓ 0.0118

 

Recall

0.8952 ∓ 0.0588

0.9107 ∓ 0.0283

 

Specificity

0.8019 ∓ 0.0567

0.9111 ∓ 0.0240

 

MCC

0.6939 ∓ 0.1095

0.8009 ∓ 0.0455

  1. Results are reported in the (average accuracy ∓ variance) format for all the 3-folds used to cross-validate the models. Note that recall is also the sensitivity metric. MCC, Matthews correlation coefficient.
Back to article page