Probit-GP classifier (Laplace approximation)

Probit-GP classifier (Laplace approximation)

Usage

morie_ghosal_np_classification(
  x,
  y,
  length_scale = NULL,
  sigma_f = 1,
  n_iter = 300,
  seed = 0
)

Arguments

x

Numeric matrix of features.

y

Numeric binary labels (0/1).

length_scale

Optional kernel length-scale.

sigma_f

Numeric signal sd (default 1).

n_iter

Integer maximum Laplace iterations (default 300).

seed

Integer RNG seed (default 0).

Value

Named list with estimate, p_hat, accuracy, length_scale, n, method.

Examples

set.seed(1)
morie_ghosal_np_classification(x = rnorm(50), y = rnorm(50))
#> $estimate
#> [1] 0.4487203
#> 
#> $p_hat
#>  [1] 0.9728062 0.2230734 0.9785746 0.9737579 0.4950818 0.1908855 0.2002538
#>  [8] 0.2257153 0.2747116 0.9189880 0.7869324 0.2042926 0.1932526 0.2261922
#> [15] 0.2253163 0.2174837 0.3831703 0.2014343 0.5118561 0.1909741 0.2330801
#> [22] 0.4665353 0.3232287 0.8156170 0.2617126 0.5685720 0.9430586 0.2073916
#> [29] 0.1900791 0.2905692 0.2317310 0.2118510 0.2163486 0.2085669 0.7589331
#> [36] 0.9572943 0.9140846 0.2165390 0.9700916 0.2271403 0.9871342 0.6182326
#> [43] 0.2154145 0.2033006 0.1927227 0.1957617 0.1915805 0.9581306 0.8338485
#> [50] 0.2327132
#> 
#> $accuracy
#> [1] 0
#> 
#> $length_scale
#> [1] 0.9016951
#> 
#> $n
#> [1] 50
#> 
#> $method
#> [1] "Probit-link GP classifier (Laplace)"
#>