Kea - Machine Learning in Marsyas ala Weka
Mode = train_evaluate
Training classifier using .arff file: mini-genres.arff
Classifier type : SVM
=== Non-Stratified cross-validation (10 folds) ===
=== ClassificationReport ===

Correctly Classified Instances		1080	83.5267 %
Incorrectly Classified Instances	213	16.4733 %
Kappa statistic				0.79408	
Mean absolute error			0.164733
Root mean squared error			0.405873
Relative absolute error			32.9466
Root relative squared error		81.1747
Total Number of Instances		1293

=== Confusion Matrix ===

	a	b	c	d	e	<-- classified as
	241	13	1	3	1	| a = classical
	25	209	0	6	19	| b = jazz
	1	0	238	16	3	| c = metal
	12	17	13	193	23	| d = pop
	3	32	3	22	199	| e = rock
83% classified correctly (1080/1293)
