pls,
If it is not a problem, can you give me more input on your use of stochastics? I've heard this term bandied about among the TA crowd, and I suspect it to be some type of indicator.
As a System Engineer, I am familiar with "stochastic" simulation and modeling. If you are trying to model an operation or system, you could use fixed data to produce a result or you could use variable data. For example, a truck is scheduled to arrive at the loading dock with raw materials at 8am, after which some operation can commence. In reality, the truck sometimes arrives at 8:10, sometimes 8:20, once in awhile as late as 8:30. But rarely is it early. You can generate a distribution curve using a sampling of arrival times.
Presumably, the operation that you are trying to simulate also has several other jobs that seem to take a variable amount of time from day to day. If you produce a simulation using these distribution curves, where the times are randomly selected, you can get a better feel for actual operational problems than if you simply used fixed average times. Then if you run several hundred simulations on your computer, you will discover that there will be days where if the the truck is 30 minutes late and such and such machine breaks down, then a potential disaster awaits. By using stochastic simulations, you can predict problems before they are allowed to happen. Whereas using just fixed data, you are only modeling a "typical" day.
Warp |
