Comparing Alternative system Configurations - CI's for the difference between the expected responses of two systems

6 important questions on Comparing Alternative system Configurations - CI's for the difference between the expected responses of two systems

Why can allowing positive correlation between X1j and X2j be of great importance?

This leads to a reduction in Var(Zj) and thus to a smaller CI.

Modified Two-sample-t CI / Welch CI

Assumption: X1j and X2j must be independent
CI: X1bar(n1) - X2bar(n2) +- t(fhat,1-α/2) sqrt(S1^2(n1)/n1 + S2^2(n2)/n2))

*modified because unequal variances

For what can the modified two-sample-t CI be used?

- The difference between the expected responses of two systems.
- For the validation of the real system observations versus simulated data.
  • Higher grades + faster learning
  • Never study anything twice
  • 100% sure, 100% understanding
Discover Study Smart

Why can we use CRN (Common random number) for simulating 2 systems?

CRN can often lead to considerable reduction in Var(Zj) and thus a smaller confidence interval
(Zj is difference both system)

What is the Behrens-Fisher problem?

The problem of comparing 2 systems with unequal and unknown variances

Which problem can be occur by comparing steady-state simulations and how to solve?

A problem occur if we simply replicate the model, because the initilization effects may bias the output.
Solution: Replication/deletion can adapted to the problem of constructing a CI for the differene between 2 steady-state means

The question on the page originate from the summary of the following study material:

  • A unique study and practice tool
  • Never study anything twice again
  • Get the grades you hope for
  • 100% sure, 100% understanding
Remember faster, study better. Scientifically proven.
Trustpilot Logo