Solving crime – and more – with principal component analysis

Imagine you're an investigator sorting through 6,000 mug shots taken at different jails around the state. You are tasked with finding all the mug shots of a high-profile con artist who has been arrested in multiple cities using a number of different aliases over the last five years.

How can you use math to help? With principal component analysis.

Principal component analysis of facial photos

The principal components of your mug shots might look like this.

Principal component analysis is a statistical technique that groups multiple variables together into several components, to reduce the amount of data that needs to be analyzed.

Using the mug shots in our example, the process identifies a selection of correlated measurements, maybe the distance between the eyes, the width of the nose and the height of the ears. It then combines those measurements into a reduced number of newly created features – the principal components – that best represent the variation in the images. Finally, the computer compares your suspects’ unique features to the principle components and quickly finds five mug shots that match.

This simple example illustrates a complex process that is used in many industries to solve problems like: Read More »

Post a Comment

Oil portfolio restructured and back in the money

Though crude oil prices edged up last week, the market remains well below VirtualOil’s original $50 strike price, meaning the hypothetical portfolio’s production is shut in in the spot market again. Oversupply continues while China GDP forecasted growth is slowing. Given the market outlook, the VirtualOil board has decided to fold the company and, with a hypothetical capital injection, restructure its portfolio at a lower strike price.

Our fictitious options-based VirtualOil portfolio launched just as the market was feeling the initial impacts of the drastic decline of oil prices. We started trading at $75 and today the market is struggling to maintain $45. That $30 of lost value on notional crude reserves of 10 million barrels/day equates to significant losses.

The decision to restructure is an appropriate reflection of some of the hard decisions many producers have faced in this continued low-price environment. Considering the M&A activity taking place in the oil patch, VirtualOil winding down its original structure and reorganizing at a lower strike price reflects some of the consolidation we see in the industry.

In the year since we ramped up the simulation, VirtualOil has netted about $15 million. Closing out the remainder of our positions, we were able to unwind an additional $35 million in value from the portfolio. We then went back to our virtual investors and raised an additional $150 million to restructure the company at a $25 strike price.

Read More »

Post a Comment

Augmented reality: yes or no?

“Our job at the lab is to find technology that’s new, cutting edge and cool – and to give it to someone else to be brilliant with.” – Matthew Horn, Manager, SAS Emerging Technologies UI Lab

Matthew Horn assists John Anderson, Senior Software Developer, with augmented reality glasses from Epson, the Moverio BT-200.

Matthew Horn assists John Anderson, Senior Software Developer, with augmented reality glasses from Epson, the Moverio BT-200.

If you’ve heard of Google Glasses or Occulus Rift, you’ve heard of augmented reality. But how does augmented reality differ from virtual reality? And how might you realistically use it in a business setting?

I spoke with Matthew Horn, Manager of the SAS Emerging Technology UI Lab to learn more about this technology and its usefulness for business. Keep reading to learn from what he told me.

What it is
Matthew Horn: Augmented reality puts data on top of what you normally see.

What it isn’t
Horn: Augmented reality is not just a head’s up display, and it’s not the same thing as virtual reality. Head’s up displays provide information on a dashboard or screen that may or may not be related to what you’re currently looking at. The speedometer on your dashboard is an example. Virtual reality, on the other hand, replaces everything you see. It immerses you in a computer generated world and completely replaces your vision. Augmented reality is a nice blend of the two. You view the world through glasses or goggles that react to what you’re already looking at by laying relevant information on top of what you’re seeing. Read More »

Post a Comment

"Five Eyes" analytics experts talk intelligence, data sharing

Cybersecurity, intelligence and analytics will also be key topics at next week’s SAS Government Leadership Forum.

Cybersecurity, intelligence & analytics will also be hot topics at the SAS Government Leadership Forum.

Following World War II, an intelligence-sharing partnership was formed between Australia, Canada, New Zealand, the United Kingdom and the United State of America.

The original alliance, called Five Eyes, was focused on monitoring the communications of the major global threat at the time – the Soviet Union. As the world has developed, so has the focus of the Five Eyes cooperative.

Read More »

Post a Comment

There’s no Internet of Things without event stream processing

The Internet of Things (IoT) is a revolutionary approach that will radically change our lives, our way of integrating with technology, and the way we do business and marketing. Companies have already defined a strategic plan for collecting and organizing data coming from the IoT. The next step is to align it with business strategy and create value from this data using existing approaches and new tools, like event stream processing.

According to Gartner, there are nearly 5 billion connected devices throughout the world today, and they predict there will be more than 25 billion by 2020.

There are nearly 5 billion connected devices throughout the world today -- and will be more than 25 billion by 2020.

Even though we have just taken the first step in the IoT revolution - with objects such as Nest, WeMo, Spiroscout and Nike FuelBand - this evolution will occur over a very short space of time. In a few years, we’ll all be diving into connected devices.

According to Gartner, there are nearly 5 billion connected devices throughout the world today, and they predict there will be more than 25 billion by 2020. Furthermore, Gartner estimates that IoT technology will be the standard for objects that we use in everyday life – we’ll have smart cars, smart washing machines, smart lighting and heating and more. Read More »

Post a Comment

How and why education customers are using SAS

Recently, I interviewed four higher education customers to hear firsthand how each is using analytics in real life (IRL). In this blog series, we will learn how educators are using analytics, why they chose SAS, and the impact it has had on their users and their institutions. In addition, they will share best practices for analytics and tips for gaining buy-in for reporting and analytics projects.

I want to thank each of the following wonderful customers for taking time to chat with me and sharing their insights on these topics.

  • Gina Huff, Senior Applications Programmer Analyst at Western Kentucky University
  • Karl Konsdorf, Acting Director, Research, Analytics and Reporting at Sinclair Community College
  • Dan Miller, Director for Business Intelligence for the North Carolina Community College System
  • Sivakumar Jaganathan, Executive Director, Data Warehouse and Business Analytics for the University of Connecticut

In this first post of the blog series, let’s start with learning how each is using SAS and why they chose SAS. Read More »

Post a Comment

Hey, hey mama, it ain’t easy being a black dog

Black Dog Syndrome may sound like the clinical term for a Led Zeppelin earworm but, as any animal shelter worker will tell you, it actually describes the challenge black dogs face in getting adopted. This phenomenon is in the spotlight today because Oct. 1 is National Black Dog Day!

This is Rocky, the adorable dog of SAS Voices editor, Alison Bolen. How could you not adopt that??

This is Rocky, the adorable dog of SAS Voices editor, Alison Bolen. How could you not adopt that??

Black Dog Syndrome was explored by 10 year-old Carah Gilmore, who used data visualization software to analyze and present data that would help a charity dear to her heart: a local pet adoption agency. Carah uncovered factors that determine how fast a dog gets adopted. Her analysis verified Black Dog Syndrome, the idea that black dogs spend more time waiting for a new home than lighter-colored dogs.

Carah and two other budding data scientists used analytics in their fifth grade science projects to learn more about their passions. Ryan Chase analyzed World Cup data. Jane Phillips created her own “Big Data” experiment. This caught the eye of people at analytics software provider SAS, who invited the kids and their parents to the company’s world headquarters for a special day.

Jane Phillips designed an experiment to engage visitors to help her build out a data set. She set up a mini basketball court and challenged visitors to sink two baskets from a distance of about 8 feet. She recorded the number of attempts each visitor needed to complete the task, and recorded the results visually. Over two days, results from over 100 visitors were recorded. She said, “By the end, I could pretty much predict in my head how many shots it would take someone based on what I had already seen. Pretty cool.”

Read More »

Post a Comment

Three tips for building a data scientist team

Maths GeniusIf I were to believe the feedback I get, statisticians are among the most difficult people to work with. What’s more, they’re the only group that should be allowed to work in data analytics. It sounds harsh, but this may explain why big data projects continually fail.

Businesses need statisticians who are both easy to work with and can take the conversation beyond math and statistics to actual business solutions. Because, not surprisingly, conversations based primarily on maths and statistics do not solve business problems -- far from it.

Businesses need to overcome the perception that data science is about feeding data into an engine and analyzing statistics to get answers. Answers require a logical mind as well as a creative one. And the starting point should be: What are the business challenges we need to solve? The statistics bit comes later and is just part of the process in getting to your business solution. Read More »

Post a Comment

Technologies for the future: yes or no

When I started college 25 years ago, we didn’t use email. I moved into the dorms my freshman year with a Brother Word Processor, convinced I would never have a single computing need beyond the necessity to type, save and print text. It’s incredible to consider how wrong I was. I never saw the Internet coming, let alone smart phones or the Interet of Things.

Matthew Horn looks at three monitors plus an iPad

Matthew Horn in the SAS Emerging Technologies UI Lab

Unlike me, Matthew Horn, who manages the SAS Emerging Technologies UI Lab, was dismantling PCs and writing computer programs around that same time period. Still, he never predicted his work with computers would go from punch cards to optical discs.

Which brings us to the next 25 years. In 2040, will you be viewing analytics dashboards on a transparent plasma screen, visualizing data overlays through augmented reality glasses or even perceiving data outputs right inside your brain? Maybe you’ll receive alerts from an implanted ear piece. Or maybe all analytic based decisions will become operational, transmitting triggers from machine to machine without the need for human intervention.

Read More »

Post a Comment

Your brain versus analytics

cyber_brain_1MBWhen I was discussing decision making and analytics with a colleague, he recommended I read the book Your Brain at Work by David Rock. I took his advice because I wanted to find out how the brain processes information and how it might relate to analytics.

Rock (you gotta love that name) explains the importance of the prefrontal cortex and how it sorts information during the decision making process. He says the prefrontal cortext is like a stage and there are a limited number of actors (information/thought processes) allowed on the stage at one time.

Read More »

Post a Comment