Fall 2018 Webinars
We are pleased to offer the following series of webinars (all times PDT). Each webinar is 60 minutes long.
- Introduction to Probability Management (Sam Savage): Monday, October 1, 10:00 AM
- SIPmath Modeler Tools Basics (Brian Putt): Tuesday, September 18, 9:00 AM and Thursday, October 18, 4:00 PM
- Interactive Decision Support at Lockheed Martin Using Probability Management Concepts (Phil Fahringer): Wednesday, November 7, 9:00 AM
- The Metalog Distributions (Tom Keelin): Wednesday, September 26, 9:00 AM
Webinars are $50 each or $150 for all webinars. To register for all webinars: 1) select "Multi-webinar package" for one webinar and complete registration; 2) Contact Mary Claire Meijer for the discount code to waive registration fees for the other webinars prior to registration. Sponsors may attend at no charge and should contact Mary Claire Meijer or Melissa Kirmse for the discount code.
Monday, October 1, 10:00 AM PDT
Introduction to Probability Management
The discipline of probability management is a revolutionary approach for communicating uncertainty as data. It may be used with everyday software such as Excel, Matlab, and R. Unlike traditional techniques, it allows simulation models to be rolled up into consolidated risk statements. Topics include:
1. The Flaw of Averages: Learn why expressing uncertainties as a single “average” numbers causes projects to be behind schedule, beyond budget, and below projection. Learn how to make consistently better decisions by explicitly modeling uncertainty.
2. Communicating Uncertainty as Data: The discipline of probability management communicates uncertainty as arrays a data called SIPs (Stochastic Information Packets). SIPs may be communicated across platforms across the enterprise to generate consolidated risk statements.
3. SIPmath, The Arithmetic of Uncertainty: Performing calculations with SIPs is called SIPmath. In native Excel, for example, you may add, multiply or use any other Excel formula on uncertainties using the same keystrokes you would have used for numbers. The magic is the Excel Data Table, which can perform thousands of simulation trials per keystroke. The free SIPmath tools available at our website make it easy to create such simulations, but the resulting models run in native Excel with no macros or add ins.
4. Real World Examples: Examples of actual applications will be demonstrated including:
● Portfolios of R&D Projects
● Rolling up Operational Risk
Tuesday, September 18, 9:00 AM PDT
Thursday, October 18, 4:00 PM PDT
SIPmath Modeler Tools Basics
In this tutorial webinar, you will learn how to use the free 3.0 SIPmath Modeler Tools, available on our website. With these tools you can easily create dynamic simulation models that can run in Microsoft Excel without any macros or add-ins. Create new models or make your existing deterministic model robust to reflect uncertainty using the SIPmath Tools.
New Features in 3.0 Version:
- Generates a wide array of probability distributions to include the Myerson and Metalog distributions.
- Utilize and control separate random number seeds to ensure repeatability in the model.
- Easily calculate Means and percentiles of random variables.
- Generates histograms of the distribution and Cumulative Distribution Graphs that are automatically updated with model changes.
- Supports Excel 2010 and newer in Windows, Excel 2016 and newer on Mac.
Wednesday, November 7, 9:00 AM PDT
Interactive Decision Support at Lockheed Martin using Probability Management Concepts
Philip Fahringer, Lockheed Martin Fellow, will provide a live demonstration of Decision Support Applications implemented at Lockheed Martin using Probability Management approaches and concepts. These applications have directly led to improved customer and business outcomes through better communication of analytic insights to help better balance performance and risk and inform decision making.
Wednesday, September 26, 9:00 AM PDT
The Metalog Distributions
The metalog distributions are an innovative new family of continuous probability distributions that better meet many of today’s needs than conventional distributions from centuries past. The metalogs can represent a much wider range of shapes than conventional distributions like the normal, lognormal, beta, or triangular. Being quantile-parameterized, metalogs automatically mold themselves to assessed or empirical data, eliminating any need for curve-fitting. They offer a choice among unbounded, semi-bounded, and bounded forms and have simple, closed-form, easy-to-program equations – making them ideal for decision analysis, simulation, and instant representation of most any assessed or empirical probabilistic data.