Skip to contents

lam * dense + (1 - lam) * sparse, ranked; ranking is 0-BASED.

Usage

morie_kamath_hybrid_retrieval_fusion(s_dense, s_sparse, lam, normalize = FALSE)

Arguments

s_dense, s_sparse

Arm scores.

lam

Mixing weight in [0, 1].

normalize

Min-max the arms first.

Value

List with scores, ranking, estimate.

Examples

morie_kamath_hybrid_retrieval_fusion(c(0.1, 0.9), c(0.8, 0.2),
    0.25)
#> $scores
#> [1] 0.625 0.375
#> 
#> $ranking
#> [1] 0 1
#> 
#> $estimate
#> [1] 0.625
#> 
#> $lam
#> [1] 0.25
#> 
#> $normalized
#> [1] FALSE
#> 
#> $n
#> [1] 2
#> 
#> $method
#> [1] "Hybrid dense/sparse score fusion"
#>