Stress testing incorporating causal dependency [10]

Go to: Summary | Previous | Next   
Bullet points include: Could build up a multi-way associative probability table But it is a lot easier (and more reliable?) to build a (much simpler) model that includes causal dependencies, e.g. Season of year affects probability of rain and of sprinkler being on or off (but not vice-versa) Status of sprinkler (on or off) and whether it has rained affects probability of pavement being wet (but not vice versa) Wetness of pavement affects its slipperiness (but not vice-versa)

Contents | Prev | Next | ERM Lecture Series

Desktop view | Switch to Mobile