Skip to contents

Mean squared error between projected features and tau_X(t), minimised over candidates. argmin is 0-BASED.

Usage

morie_kamath_ch9_output_projector_mse(H_X, tau_X, t)

Arguments

H_X

A feature matrix, or a list of candidates.

tau_X

Target features or a function of t.

t

Target text.

Value

List with estimate, argmin, losses.

Examples

morie_kamath_ch9_output_projector_mse(H_X = c(1, 2, 3, 4, 5, 6, 7, 8),
  tau_X = c(1, 2, 3, 4, 5, 6, 7, 8), t = c(1, 2, 3, 4, 5, 6, 7, 8))
#> $estimate
#> [1] 0
#> 
#> $argmin
#> [1] 0
#> 
#> $losses
#> [1] 0
#> 
#> $n_candidates
#> [1] 1
#> 
#> $n
#> [1] 8
#> 
#> $method
#> [1] "output-projector MSE objective (Kamath Eq 9.18)"
#>