A Survey of Computational Intelligence Techniques in Protein Function Prediction

Table 10

Summary of computational intelligence techniques in pathway analysis from gene expression data.

Reference	CI techniques	Performance	Datasets
[139]	Gene set enrichment analysis	Sensitivity: 0.78, specificity: 0.98, AUC: 0.94	Gene expression data with significance analysis of microarray
[140]	Linear discriminant analysis	Error rate: 10–15%	Covariance matrix with group relationships among variables
[141]	Random forest	Error rate: 11–17%	Gene expression data
[142]	Naïve Bayes, decision tree based ensemble classifier	Accuracy: 91.2% and -measure: 0.787	Gene expression data
[143]	SVM, Bayesian approach, C5.0, and random forest	Error rate: 7–15%	Gene expression data
[144]	Bayesian approach	AUC: 90.56%, Accuracy: 75.7%	Single-nucleotide polymorphisms