Table 1.

The effect of the choice of features on spectral clustering

Optimal feature set								${\mathit{\mu }}_1$	${\mathit{\mu }}_2$	${\mathit{\mu }}_{sc}$
${\mathit{\xi }}_2$								0.459	0.225	−0.234
${\mathit{\xi }}_2$	${\mathit{\xi }}_8$							0.514	0.197	−0.317
${\mathit{\xi }}_2$	${\mathit{\xi }}_8$	${\mathit{\xi }}_7$						0.491	0.205	−0.286
${\mathit{\xi }}_2$	${\mathit{\xi }}_8$	${\mathit{\xi }}_7$	${\mathit{\xi }}_5$					0.521	0.193	−0.327*
${\mathit{\xi }}_2$	${\mathit{\xi }}_8$	${\mathit{\xi }}_7$	${\mathit{\xi }}_5$	${\mathit{\xi }}_1$				0.495	0.206	−0.286
${\mathit{\xi }}_2$	${\mathit{\xi }}_8$	${\mathit{\xi }}_7$	${\mathit{\xi }}_5$	${\mathit{\xi }}_1$	${\mathit{\xi }}_4$			0.492	0.235	−0.257
${\mathit{\xi }}_2$	${\mathit{\xi }}_8$	${\mathit{\xi }}_7$	${\mathit{\xi }}_5$	${\mathit{\xi }}_1$	${\mathit{\xi }}_4$	${\mathit{\xi }}_3$		0.483	0.235	−0.248
${\mathit{\xi }}_2$	${\mathit{\xi }}_8$	${\mathit{\xi }}_7$	${\mathit{\xi }}_5$	${\mathit{\xi }}_1$	${\mathit{\xi }}_4$	${\mathit{\xi }}_3$	${\mathit{\xi }}_6$	0.450	0.243	−0.207

Notation (${}^{\ast }$) points out on the selected feature subset, ${\mathit{\Xi }}^{\ast },$ that reaches the lowest value of ${\mathit{\mu }}_{sc}$