Application of Legendre polynomials based neural networks for the analysis of heat and mass transfer of a non-Newtonian fluid in a porous channel

Advances in Continuous and Discrete Models

Theory and Modern Applications

Table 2 Parameter setting for Generalized normal-distribution algorithm and sequential quadratic programing

Algorithm	Parameters	Settings	Parameters	Settings
LeNN-GNDO	Technique	Metaheuristic	Candidate selection	Random search
	Max. Iterations	5000	Function tolerance	10⁻¹⁸
	Bounds (Lb, Ub)	[−1,1]	Fitness Limit	10⁻¹⁵
	Search agents	70	Other settings	Default
SQP	Initial weights	Best of LeNN-GNDO	Function evaluations	200,000
	X-Tolerance ‘TolX’	1.00E−20	Max iterations	3000
	Fitness Limit	10⁻¹⁵	Other settings	Default