Table 3.

Results of classification of surface defects. LDA: linear discriminant analysis; knn: k-nearest neighbors; QDA: quadratic discriminant analysis; SVM (rbf): SVM with radial basis function kernel; SVM (poly): SVM with polynomial kernel function; SVM (quad): SVM with a quadratic kernel.

Classifier	knn (k = 10)	LDA	QDA	Multilayer Perceptron	SVM (rbf)	SVM (poly)	SVM (quad)
Accuracy	90.09%	86.22%	66.89%	93.95%	94.17%	87.72%	93.05%