Predicting the likely behaviors of complex systems

The author discusses a method to predict the likely behaviors of a complex system when one has a model of that system, but not all the inputs/parameters are precisely known. Depending on the given model, the method can answer such questions as how likely it is that some state variable(s) will stay inside (outside) of a certain numeric range, or increase (decrease); the system will go into oscillation vs. remaining stable; any transients will be below a given threshold and last less than a given duration. Applications include predicting semiconductor yields and drug safety and effectiveness. The author is concentrating on using a cardiovascular model to predict the possible effects of heat disease therapies.<<ETX>>