ICEAA TechShowcase: Modeling OLS-based CER Errors

Palisade Presents:

Modeling OLS-based CER Errors

If your cost model includes an Ordinary Least Squares (OLS)-based Cost Estimating Relationship (CER) and you are not explicitly modeling the associated error, for a linear regression you are implicitly assuming that there is no error.  On average that’s a safe assumption; it’s individual occurrences however that can make this risky.  After a brief review of OLS regression fundamentals (yeah, calculus), this webinar will show how @RISK can be used to model a regression equation’s error term for both linear and nonlinear (intrinsically linear) CERs.  Along the way you will improve your skill working with the normal distribution and, if it’s not already part of it, add the lognormal distribution to your repertoire.

Presented by Dr. Steve Van Drew

Dr. Steve Van Drew embodies the jack-of-all trades lifestyle as a former leader/manager, academician, and analytical practitioner, with a multi-disciplinary education in engineering, applied economics, and management science/operations research. Steve enjoys a busy “retirement” near Richmond, VA, and yet manages to dedicate time to share his passion for spreadsheet modeling and @RISK.