Table 3.

Comparison of the performance between PHTNet and three well-regarded PHT processing frameworks.

Dataset		BMFLC-PHTNet	EBMFLC-PHTNet	WAKE-PHTNet
Validation	p-value	$5.96E-08$	$5.96E-08$	$5.96E-08$
	Improvement	$+\mathrm{88.42\%}$	$+\mathrm{87.78\%}$	$+\mathrm{90.18\%}$
Test	p-value	$5.96E-08$	$5.96E-08$	$5.96E-08$
	Improvement	$+\mathrm{89.58\%}$	$+\mathrm{89.01\%}$	$+\mathrm{91.07\%}$

The $p-values$ are derived based on 95% confidence intervals. The numbers in the “Improvement” rows represent the improvement in the mean Normalized MSE obtained when moving from other techniques to the PHTNet.