View as a web page

SAS® Statistics and
Operations Research News
March 2015  |  subscribe  |  unsubscribe
youtube blogs sascommunity
Springing into SAS® Global Forum 2015

The excessive winter in parts of the United States seems to be abating, and the fact that we are now on eastern daylight savings time here at SAS headquarters points to spring and SAS Global Forum. So I have to finish all my materials for the conference sooner rather than later! Here's hoping that readers presenting at the conference are also finishing up in good stead, and that those attendees from the Boston area can shovel themselves out in time to get down to Dallas!

As always, this edition highlights some of the related analytical content at the upcoming conference, and those readers not going can expect to benefit from the windfall of papers and online videos that will follow. This year, the lineup looks especially appealing to statisticians and data scientists. In particular, Jennifer Waller, Pre- and Post-Conference Coordinator and next year’s conference chair, is excited about the analytics tutorial lineup: “Not only have we expanded the SAS statistical tutorials into Sunday afternoon again this year, we’re also pleased with the external presenter lineup, including tutorials on missing data analysis in clinical trials by Michael O’Kelly and Sonia Davis, forecasting by David Dickey, and survival analysis by Joseph Gardiner.”

O'Kelly and Davis taught their missing data course, based on the book Clinical Trials with Missing Data (O'Kelly and Ratitch), at a conference here last spring and it was a big success. Joseph Gardiner and Dave Dickey are renowned teachers and their students are in for a treat.

I am also looking forward to the keynote address by Alan Schwarz, an amazing reporter at the New York Times whose investigative journalism and incorporation of analytics into his work have led to a greater appreciation of the dangers of head injuries in sports.

All right. Besides pointing out the analytical content provided by SAS at SAS Global Forum, this newsletter focuses on SAS® Simulation Studio, software from the Operations Research R&D group that builds discrete-event simulation models. See later in the newsletter for an introduction, a description of an application for managing staffing at a neonatal intensive care unit, and a new video that shows SAS Simulation Studio in action. In addition, we highlight a "film series" starring SAS/ETS® software and present various tips and other information that will hopefully make your analytical life a little easier.

To spring!


Senior R&D Director, Statistical Applications

Featured Papers
Creating Path Diagrams That Impress »

In structural equation modeling, researchers often use path diagrams to represent their models graphically. Path diagrams enable you to visualize the conceptual models behind the research and to depict statistical results in an intuitive way. In SAS/STAT® 13.1, the CALIS procedure begins to produce high-quality graphical output of path diagrams from model specifications, with default diagrams as well as a facility for customization. Examples illustrate how to produce path diagrams that have different emphases, styles, or formats. This paper also demonstrates the use of the ODS Graphics Editor to edit path diagrams that are produced from the CALIS procedure.

Penalized Regression Methods for Linear Models in SAS/STAT »

Regression problems with many potential candidate predictor variables occur in a wide variety of scientific fields and business applications. These problems require you to perform statistical model selection to find an optimal model, one that is as simple as possible while still providing good predictive performance. Traditional stepwise selection methods, such as forward and backward selection, suffer from high variability and low prediction accuracy, especially when there are many predictor variables or correlated predictor variables (or both). In the last decade, the higher prediction accuracy and computational efficiency of penalized regression methods have made them an attractive alternative to
traditional selection methods. This new paper first provides a brief review of the LASSO, adaptive LASSO, and elastic net penalized model selection methods. Then it explains how to perform model selection by applying these techniques with the GLMSELECT procedure, which includes extensive customization options and powerful graphs for steering statistical model selection.

Simulating Portfolio Losses from Adverse Events: Applications to the Insurance and Finance Industries »

Companies in the insurance and banking industries need to model the frequency and severity of adverse events every day. Accurate modeling of risks and the application of predictive methods ensure the liquidity and financial health of portfolios. Often the modeling involves computationally intensive, large-scale simulation. SAS/ETS provides high-performance procedures to assist in this modeling. This paper discusses the capabilities of the HPCOUNTREG and HPSEVERITY procedures, which estimate count and loss distribution models in a massively parallel processing environment. The loss modeling features have been extended by the new HPCDM procedure, which simulates the probability distribution of the aggregate loss by compounding the count and severity distribution models. PROC HPCDM also analyzes the impact of various future scenarios and parameter uncertainty on the distribution of the aggregate loss. This paper steps through the entire modeling and simulation process that is useful in the insurance and banking industries.

Introduction to SAS Simulation Studio »

An overview is presented of SAS Simulation Studio, an object-oriented, Java-based application for building and analyzing discrete-event simulation models. Emphasis is given to Simulation Studio's hierarchical, entity-based approach to resource modeling, which facilitates the creation of realistic simulation models for systems with complicated resource requirements, such as preemption. Also discussed are the various ways that Simulation Studio integrates with SAS® and JMP® for data management, distribution fitting, and experimental design.

See SAS Simulation Studio in action for exploring system performance.

Creating a SimNICU: Using SAS Simulation Studio to Model Staffing Needs in Clinical Environments »

Patient safety in a neonatal intensive care unit (NICU)—as in any hospital unit—is critically dependent on appropriate staffing. This project used SAS Simulation Studio to create a discrete-event simulation model of a specific NICU that can be used to predict the number of nurses needed per shift. The model incorporates the complexities inherent in determining staffing needs, including variations in patient acuity, referral patterns, and length of stay. To build the model, the group first estimated probability distributions for the number and type of patients admitted each day to the unit. Using both internal and published data, the team estimated distributions for various NICU-specific patient morbidities, including type and timing of each morbidity event and its temporal effect on a patient's acuity. The final simulation model samples from these input distributions and simulates the flow of individual patients through the NICU (consisting of critical-care and step-down beds) over a one-year period. The general basis of the model represents a method that can be applied to any unit in any hospital, thereby providing clinicians and administrators with a tool to rigorously and quantitatively support staffing decisions. With additional refinements, the use of such a model over time can provide significant benefits in both patient safety and operational efficiency.

SAS Analytical Presentations at SAS Global Forum 2015 »

If you're heading to SAS Global Forum, take a look at the many offerings by the SAS analytical staff. They include paper presentations, tutorials, hands-on workshops, and super demo presentations in the exhibition hall, where you should also come by and talk with the development staff about what they are doing (and what you'd like them to do!). And this is just the tip of the iceberg, as the analytical presentations and tutorials by our users look outstanding. If you aren't attending, there's still plenty of time to register!

A Note about SAS® University Edition for You Teachers and Learners »

If you download the current version of SAS University Edition, you get the most recent versions of SAS/STAT and SAS/IML® software, which are the 13.2 releases. If you downloaded the SAS University Edition before October 2014, and you use the Update feature, you will get updates for that release only. In that case, you will need to download the latest version and replace your original version (or keep them both, which is easy to do with VMware). Just run proc product_status; to find out what you have. For more information, see the SAS University Edition Help Center.

The DO Loop »

Several recent posts from Rick Wicklin’s blog bear pointing out. He discusses how to approximate a cumulative distribution function (CDF)  as well as how to create a custom probability density function (PDF) and CDF by using SAS. Wicklin also discusses creating arrays of matrices.

Lights, Camera, Action! »

This newsletter’s film series focuses on SAS/ETS software. Learn about using the TIMEDATA procedure for preparing your time series data; SAS/ETS tools for modeling frequency, severity, and aggregate loss data; and modeling aggregate losses to estimate capital requirements for risk management. Grab some popcorn and enjoy!

Tech Support Points Out
Modeling rates and estimating rates and rate ratios (with confidence intervals) »

When the count of an event is observed over a period or amount of exposure, such as deaths per 100,000 individuals, traffic accidents per year, or injuries per person-year, it is called a rate. Unlike a proportion, which ranges from 0 to 1, a rate can have any nonnegative value, such as 4.2 deaths per 100,000 individuals or 65 accidents per year. You often use a Poission or negative binomial model to fit models to such data. This note illustrates rate and rate ratio estimation in a Poisson model or negative binomial model.

Talks and Tutorials
SAS Global Forum 2015 »

April 26-29, 2015

Dallas, TX

Sessions at a glance

Pre-Conference Tutorials

Featured Talks and Presentations

Visit with SAS at the INFORMS Conference on Business Analytics and OR, April 12–14 »

In mid-April, SAS staff will be in Huntington Beach, California, to attend and present at this annual gathering of analytics practitioners. The conference this year is chaired by Manoj Chari, who directs the OR R&D Department at SAS. Brad Klenz will give a talk titled "Event Stream Processing for Power Grid Analysis," and Jeff Day will present a case study to the Professional Colloquium for students. Polly Mitchell-Guthrie, chair of the INFORMS Analytics Certification Board, will speak in a session on certification. Stop by the SAS booth anytime to visit, ask questions, or see our latest features.

Statistics and Operations Research Home »
SAS/STAT Ambassador Kit »
Bayesian Resources »
SAS/STAT® 13.1 and 13.2 Resources »
SAS/STAT® 9.22 Resources »
FASTats: Frequently Asked-For Statistics »
SAS Discussion Forums »
Software Product Pages A-Z »
SAS/STAT® Procedures A-Z »
Getting Started Resources for SAS/STAT 9.2 »
Analytical SAS® Software Video Portal »
SAS/STAT 13.2 Example Programs (Sample Library) »
SAS/STAT 13.1 Example Programs (Sample Library) »
SAS/ETS 13.2 Example Programs (Sample Library) »
SAS/ETS 13.1 Example Programs (Sample Library) »
Technical Support
E-newsletter Archives »
Technical Problems »
Don't miss important updates from SAS! Please add as a domain
in your safe sender list.
Your Subscription
We hope that you enjoyed reading this e-newsletter. However, if you would rather not receive SAS e-newsletters in the future, return to our e-newsletter subscriptions site to log in and edit the subscription options in your profile.
Tech Support | Privacy Statement

To contact SAS via postal mail: SAS, SAS Campus Drive, Cary NC 27513 USA. ATTN: Legal Division/Privacy Manager.

SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks
of SAS Institute Inc. in the USA and other countries. ® indicates USA registration. Other brand and
product names are trademarks of their respective companies.

Copyright © SAS Institute Inc. All rights reserved.