Table 5.

Classification results of seven features based on different values of k and distance metrics for Dataset I (%).

Distance	k	MV	FFT	Db1	Db2	Db3	Db4	Db5
EU	1	68.75	73.75	90.00	91.25	87.50	88.75	83.75
	3	63.75	72.50	75.00	78.75	78.75	81.25	80.00
	5	46.25	45.00	66.25	70.00	70.00	71.25	73.75
	7	51.25	53.75	60.00	68.75	70.00	73.75	75.00
	9	43.75	60.00	57.50	66.25	66.25	66.25	70.00
CB	1	66.25	70.00	91.25	91.25	86.25	86.25	82.50
	3	60.00	62.50	71.25	75.00	76.25	77.50	75.00
	5	41.25	47.50	61.25	71.25	70.00	72.50	72.50
	7	52.50	53.75	57.50	71.25	71.25	72.50	73.75
	9	48.75	48.75	55.00	65.00	63.75	65.00	67.50
COS	1	77.50	77.50	90.00	92.50	91.25	91.25	86.25
	3	72.50	78.75	80.00	82.50	82.50	82.50	82.50
	5	57.50	60.00	68.75	68.75	72.50	73.75	80.00
	7	58.75	53.75	63.75	65.00	67.50	71.25	75.00
	9	46.25	43.75	55.00	62.50	61.25	61.25	66.25
COR	1	78.75	77.50	93.75	92.50	91.25	92.50	87.50
	3	71.25	76.25	81.25	85.00	85.00	85.00	85.00
	5	56.25	57.50	66.25	76.25	75.00	76.25	82.50
	7	52.50	61.25	63.75	66.25	67.50	71.25	76.25
	9	51.25	58.75	50.00	58.75	63.75	65.00	67.50

Note: EU, Euclidean distance; CB, cityblock distance; COS, cosine distance; COR, correlation distance. The bold numbers are the highest accuracies for each distance metric.