Documentation for "Explain Predictions"?

sherlock
sherlock New Altair Community Member
edited November 2024 in Community Q&A
Hello,

seem "Explain Predictions" Operator gives a nice balance of calculation time and level of explainabiliity. For scientific use and some customers
it would be required to explain the algorithm behind. Seems to be close to LIME, but e.g. how exactly does it calculate global feature importance?

Thanks!
Tagged:

Categories