Table 5.

Estimated parameters of the mixture of up to six Gaussian functions and the Gumbel distribution using a maximum-likelihood approach with the computed log-likelihood values

Mixture of k Gaussian distributions
k	Estimated parameters	Log-likelihood
1	w = 1	−13,018
	μ = 0.24
	σ = 1.67
2	w = (0.86, 0.14)	−10,818
	μ = (−0.18, 2.77)
	σ = (0.80, 2.92)
3	w = (0.69, 0.27, 0.04)	−10,396
	μ = (−0.45, 1.29, 5.31)
	σ = (0.58, 0.98, 4.19)
4	w = (0.57, 0.32, 0.09, 0.02)	−10,296
	μ = (−0.61, 0.73, 3.05, 9.11)
	σ = (0.49, 0.71, 1.46, 7.46)
5	w = (0.48, 0.34, 0.14, 0.03, 0.01)	−10,239
	μ = (−0.73, 0.33, 1.85, 4.86, 12.29)
	σ = (0.41, 0.50, 0.78, 1.69, 8.22)
6	w = (0.39, 0.33, 0.18, 0.07, 0.02, 0.01)	−10,229
	μ = (−0.85, 0.02, 1.11, 2.68, 5.65, 14.35)
	σ = (0.36, 0.39, 0.53, 0.87, 2.08, 9.61)

Gumbel distribution
Estimated parameters	Log-likelihood
λ = −0.52; δ = 0.93	−10,723