Feature Selection Using Genetic Algorithms for the Generation of a Recognition and Classification of Children Activities Model Using Environmental Sound

Table 7

Accuracies by the generated model.

Number of features	Feature selection method	Classifier	Accuracy
5	Genetic algorithm	k-NN	0.89
5	Genetic algorithm	NC	0.815
5	Genetic algorithm	ANN	0.85
5	Genetic algorithm	RF	0.92
5	Genetic algorithm	Rpart	0.857
27	Akaike criterion	k-NN	0.9425
27	Akaike criterion	RF	0.9525
34	None	k-NN	1.00
34	None	RF	0.7637
5	Forward selection	SVM	0.7944
5	Forward selection	k-NN	0.8611
5	Forward selection	RF	0.8917
5	Forward selection	Gradient boosting	0.8710
5	Backward elimination	SVM	0.8079
5	Backward elimination	k-NN	0.8545
5	Backward elimination	RF	0.8946
5	Backward elimination	Gradient boosting	0.8797
30	Forward selection	SVM	0.8867
	Backward elimination
30	Forward selection	k-NN	0.9075
	Backward elimination
30	Forward selection	RF	0.9208
	Backward elimination
30	Forward selection	Gradient boosting	0.9228
	Backward elimination