Implementation of the algorithm proposed in the paper "ShaRP: Explaining Rankings with Shapley Values", by Pliatsika et al.
0
stars
0
forks
source link
Add possibility for parallelizing calculations and introduce `n_jobs` parameter #16
Closed
joaopfonseca closed 5 months ago
Describe the workflow you want to enable
The calculation of shapley values cannot be parallelized in the current implementation. This could be made possible in the future.
Describe your proposed solution
See
mlresearch.utils.parallel_loop
Describe alternatives you've considered, if relevant
No response
Additional context
No response