We have updated our software for improved interpretability since this post was written. For the latest on this topic, read our new series on model-agnostic interpretability. While some machine learning models – like decision trees – are transparent, the majority of models used today – like deep neural networks, random forests, gradient boosting
Artificial intelligence often seems misunderstood, especially in fraud. The same is true of machine learning. One of the amazing things about them is they ask the unasked questions. This occurs as artificial intelligence (AI) and machine learning (ML) go about their daily work. So, what is the unasked question? Too
President Donald Trump - people seem to either love him or hate him. Which makes for an interesting data-analysis topic ... tracking Trump's approval rating. Follow along as I explore some data! I'll start with what I consider the 'best' approval-rating graph I've found so far. Here's a screen-capture of
Jim Harris says people still play a pivotal role in data-driven decision making.
This article is the first in a series of three publications covering REST APIs and their use in, and with SAS. Today, I want to cover a basic example using SAS Viya REST APIs to download an image from a report in SAS Visual Analytics.
What skills will students need in order to pursue a lucrative career in analytics? I recently interviewed two professors to find out. In my first post, the professors discussed the importance of teaching and learning analytics. Today, we'll hear their top advice for students studying analytics, such as: get a
How do you deploy your model so that business processes can make use of it? This post explores how SAS Viya applications can directly add models to a model repository, and specifically focuses on how to deploy them with SAS Model Manager to Hadoop.
I consider October 31, the unofficial start of the holiday season. It begins with Halloween which brings a seemingly unlimited supply of candy. Who needs that many candy corns anyway?! Next is Thanksgiving. A one-day feast that turns into 4 days of binge watching football and eating leftovers. That’s followed
A useful feature in PROC SGPLOT is the ability to easily visualize subgroups of data. Most statements in the SGPLOT procedure support a GROUP= option that enables you to overlay plots of subgroups. When you use the GROUP= option, observations are assigned attributes (colors, line patterns, symbols, ...) that indicate
Note: Today’s utility industry is in upheaval. All of the assumptions the business has run on have been turned on their heads. This post is the second in a three-part series looking at how analytics are helping utilities navigate this challenging landscape and find new opportunities for improvements in operations,
I’m on day 8 of my 10 day sabbatical from running. By day 3, I already felt anxious about not lacing up my shoes and heading out to feel the wind in my face, never mind the fact that I was still exhausted and sore from the previous week’s races.
GAHHHHHHHHH! My screams filled the office hallways on a Sunday afternoon. In agony, I hopped to the ice machine to find relief for my crushed toe. Moments before, while hanging my shiny new patent plaque on the wall, it dropped six feet and landed on my big toe joint. Since
If you want to bootstrap the parameters in a statistical regression model, you have two primary choices. The first, case resampling, is discussed in a previous article. This article describes the second choice, which is resampling residuals (also called model-based resampling). This article shows how to implement residual resampling in
In addition to the many CAS actions supplied by SAS, as of SAS® Viya™ 3.4, you can create your own actions using CASL. Learn how in this post.
The other day I was playing around with the voter registration data for all ~8 million registered voters in North Carolina (yes - this guy knows how to have fun!), and I got to wondering what last names were the most common. I summarized the data by county, and the
How should a data trust process work? David Loshin elaborates.
“I’m so glad I live in a world where there are Octobers.” L.M. Montgomery I don’t believe I’ve ever met a person who doesn’t love October. My mom always said the sunsets in October are the best …. I agree with her! October’s crisp temps and clean air put an
If you want to bootstrap the parameters in a statistical regression model, you have two primary choices. The first is case resampling, which is also called resampling observations or resampling pairs. In case resampling, you create the bootstrap sample by randomly selecting observations (with replacement) from the original data. The
In my 15+ years working in clinical research, I've not seen significant evolution in the way we do clinical trials and apply advanced analytics to clinical trial data. A personal story demonstrates that a similar problem is at play in health care. Each year my daughter visits the children’s cardiology
Recently a colleague was requested to create a bar chart showing the performance of students in a school compared to county and state wide schools. The solution using SGPLOT was simple as shown below. /*--Create data--*/ data school; input Year Group $ Value; format Value percent5.1; datalines;
Not long ago I read this New York Times article on The Power of Positive People. After reading about the unsurprising benefits of having a constant support group, I was inspired to think of how I could create opportunities for people in my social circle to get together and support
This past Sunday was the last day of the NC State Fair, and we broke some records this year! Can you guess what those records were? I'll ramble on about the fair to give you time to come up with your guess, and then I'll let you know about the
I've been recovering for 15 years now, after a lengthy career caring for critically ill patients. Now, I’m part of a team at SAS that’s working to transform health care – and that's important to me because of something that happened when I was an ICU nurse. It changed my
Applying machine learning approaches to forecasting is an area of great research interest. Progress is being made on multiple fronts, for example: In the M4 Forecasting Competition, completed earlier this year, the top two performers utilized machine learning with traditional time series forecasting methods. At the link you'll find full
Have you noticed a lot of graphs popping up lately, about politicians, voters, and elections? ... Well of course - it's an election year! If you follow my blog posts, then you're probably a bit "smarter than the average bear" when it comes to scrutinizing graphs. And election graphs election
With SAS Data Preparation and SAS Decision Manager, you can perform out-of-the-box column and row transformations to increase your data quality and build the foundations for data-driven innovation. This blog will discuss how you can leverage SAS Decision Manager to enrich data when preparing it through SAS Data Preparation.
A SAS programmer asked how to rearrange elements of a matrix. The rearrangement he wanted was rather complicated: certain blocks of data needed to move relative to other blocks, but the values within each block were to remain unchanged. It turned out that the mathematical operation he needed is called
Did you know that you can now chat with SAS Technical Support? Technical Chat enables you to quickly engage with a knowledgeable consultant when you have a SAS question or need help with troubleshooting an issue.
Any of you who are even slightly into politics, or have followed any news lately, have probably seen something about the DNA test Elizabeth Warren took to prove/disprove her Native American ancestry. The test indicates she might have had a Native American ancestor 6 to 10 generations ago. That's a
There are four widely recognized styles of machine learning: supervised, unsupervised, semi-supervised and reinforcement learning. These styles have been discussed in great depth in the literature and are included in most introductory lectures on machine learning algorithms. As a recap, the table below summarizes these styles. For a comprehensive mapping
