Why can exhaustive adjustment-set search be costly on a large causal graph?
StatisticsLevel: IntermediateCausal identification and targets
Why can exhaustive adjustment-set search be costly on a large causal graph?
Hints
- Subset counts grow rapidly.
- Do you need every solution?
Answer
It explores many candidate variable subsets to return all valid sets. A maximal or targeted search can be faster when one valid set is sufficient.
Source and licence: DoWhy — backdoor identification · MIT · answer: Quantoros editorial