In Part 1, Udo provided SAS code to replicate the example in Hyndman's blog.  Below, he shows the results of out-of-sample testing, and draws some conclusions on the computational efficiency of this approach. Out-of-sample Testing In addition to the example shared by Hyndman, out-of-sample data was used to illustrate the final performance

 In this guest blogger post, Udo Sglavo of the Advanced Analytics Division of SAS shows how to conduct time series cross-validation using SAS Forecast Server.  Udo replicates the example from Rob J Hyndman's Research Tips blog. Replicating the Example In order to replicate the example in Hyndman's blog, the example

The hurricane didn't get me, but Monday night's thunderstorm sure played a dirty trick. After leaving my car windows cracked open over night, I drove to work Tuesday morning feeling a little soggier by the minute.  Upon arrival at SAS, I was aghast to find the seat of my pants was soaked

We had a tornado in April, an earthquake on Tuesday, a drought all summer, and a hurricane arrives on Saturday. All I can figure is that Cary, NC has way too many sinners per capita. What's next -- pestilence? The BFD Makeover The BFD (and all SAS blogs) will now be

Tuesday's release of SAS 9.3 included the new SAS Forecast Server 4.1, which has several valuable enhancements: Combination (Ensemble) Models: A combination of forecasts using different forecasting techniques can outperform forecasts produced by using any single technique. Users can combine forecasts produced by many different models using several different combination

No, BONEZONE is not the website of wayward legislators. It is, however, a trade journal of the orthopaedic devices industry, and the Summer 2011 issue features a nice mention of Forecast Value Added (FVA) analysis in an article by Tom Wallace. In "Forecasting: It's Getting Better," Tom refers to FVA

There is nothing the gambler, investor, forecaster, or Match.com dater likes as much as the sure thing. Don't we all? Back in April I stated what I claimed to be a sure thing forecast: In any group of 2 or more people, there is at least one pair of people

As organizations confront the limits of forecasting, they finally realize the folly in a blind pursuit of unachievable levels of forecast accuracy. The best accuracy we can ever hope to achieve is limited by the nature -- the forecastability -- of what we are trying to forecast. Anything better than

Have you ever wanted to look like George Clooney and get all the money, fame, and dates? I have long aspired to this. But in the great poker hand of life, I wasn't dealt an inside straight or even a nice pair. So I make do with what I've got.

The Spring 2011 issue of Foresight is now available. Here is Editor Len Tashman's preview: For forecasters, “being wrong” is the expectation; the hope is that we’re not too wrong. But admitting to our failures is never easy. The Spring 2011 issue leads off with Marcus O’Connor’s book review of