What is unicity & why you need to know

Tags

, , ,

A quick Google of “financial information breach” in the news returns an almost daily litany of public and private sector institutions that have been hacked for information.

In June 2015, the Office of Personnel Management (OPM) announced a breach that compromised over 4 million personnel records. The following month, OPM again confessed to another breach; this times it was over 21 million records, including the files used for security clearances. These specific files incorporate the background investigations which include extensive documentation of personnel employment, health and personal information. (Both instances are blamed on China.)

DON Leadership OPM Data Breach Briefing 2015-06-26

Keys to the Castle

Safeguarding personal information is a monumental task. We don’t just take it for granted that the people we give our information – health care, financial institutions, employers – will steward the data properly, we hold them accountable, both in civil and criminal court. It is easy to want an entity to be responsible and answerable to protecting personal information, but in reality, this example is only a simple liability we understand. Your personal information in reality is a much more complex picture, and infinitely more vulnerable beyond the government and corporate entities that strive to uphold you PI.

Think about your social media data stream. You probably wouldn’t be surprised that someone could figure out who you are by what you post. What does that look like and how easy is it?

The answer is …

Unicity is the a statistical tool used to measure how much “outside” information is need to identify an “anonymous” individual within a dataset. One way to measure that is how many “tuples” it takes to hit the mark. “Tuples”[1] is a “data structure — a mechanism for grouping and organizing data to make it easier to use.” Short for n-tuple or multiple in mathematics, it has n elements to set a data point. In the case of this article, that data point is your identity.

What signal are you sending?

With every purchase using a credit card, the financial transaction is specifically encrypted by sender and receiver to ensure the financial information is sufficient to protect it from hacking. That doesn’t make it anonymous though. Think Big Data.

MIT researchers analyzed a data set of more than one million people at ten thousand businesses.[2] The data was “anonymized”; whereupon, the researchers were able to see details about each transaction, such as when, where, and how much, but not allowed names or account numbers. A tuple of location and time proved a simple solution to identification. With just four of random tuples, the MIT folks was sufficient to reveal 90 percent of the individuals in the dataset.

Cash May Be King …

But it still doesn’t make you anonymous. Your data stream is not confined to credit card transactions. Every geo-tagged photo, every social media comment or use of your phone reveals who and where you are. If you want to brave turning off the virtual world to cloak your movements, you are still followed through license plate readers and shopper movements caught on camera. License plate scans are used on police cars, on bridges, roads and tollbooths to capture time and place. In brick and mortar stores, your movements, attributes and actions are captured on camera, and possibly analyzed.[3] Is that creepy? Possibly, but considering every click on Amazon or every other website on the internet is forever captured by cookies, is there a difference? Or possibly it’s only a difference you are more comfortable understanding … and feeling creepy?

Bottom line: you are rarely alone.

Traffic_Camera_Observing_M1_-_geograph.org.uk_-_765304

It’s not all about the money either.

“Life is short. Have an affair.” The Ashley Madison website terse tagline speaks terabytes of information about its content. One of many sites that provide a covert location to seek others with the same guilty intentions, Ashley Madison made the news in June 2015 as well for being hacked. It’s not the Chinese this time and it’s not ransom for money. The “Impact Team” as the hackers call themselves are demanding the website shut down in return for not releasing the financial (credit card & employment), personal (name and address), and intimate (do I have to draw a picture?) details of the site’s reported 37 million members.

Same old story?

Is this a new phenomenon? Actually, personal accountability, who and where you are and what you do, is not new. Detection, whether a picture of your car license plate or your credit card transaction, has been around for as long as cars and credit cards. Sherlock Holmes and Hercule Poirot understood data trail long before digital medium. (Well, their creators did.)

What is new is Big Data. What has changed is capability of volume, velocity and variety of information that is ubiquitously captured and shared. This aptitude used to be cost prohibitive. The total capture is now relatively inexpensive. Using the data has become a capability differentiator, let alone a potent return on investment.

The data has always been there; it’s just being used faster and funnier. That’s why you need to know unicity and the power of Big Data.

My sincere appreciation to Sherbit as inspiration and reference for this article. I highly endorse following them if you are interested in data privacy issues.

[1] http://openbookproject.net/thinkcs/python/english3e/tuples.html

[2] https://www.sherbit.io/instagram-surveillance/

[3] http://www.nytimes.com/2013/07/15/business/attention-shopper-stores-are-tracking-your-cell.html?_r=0

Five “Shoulds” for Data Driven Decision Making

Tags

,

Decision Making is one of the driving factors for data collection.  Whether it’s used to choose what route to take to work or determining Global Thermo Nuclear War strategy, data is a vital tool for interpreting options.  Many a bad decision has been made with bad information, as the New Coke campaign or the volatility of the stock market can attest.

One of the evils of statistics is using numbers to tell a biased story.  Or perhaps it’s not telling the whole story.  Or maybe the people who prepare the story don’t understand its pertinence.  Or maybe the ones who prepare the story don’t know how to tell the story well.

Whether you work dashboards or data visualization, this article by Erik K is an awesome resources for beginners and veterans on making sure your data tells the right story.

View story at Medium.com

Awesome

 

 

Professional Summer Reading on Data Analytics

Tags

, , ,

Don’t I love a great book!!  This is an awesome professional reading list for Data Analytics provided by Analytics Vidhya.   I’ve read half of them, so I’ll get cracking on the rest, starting with Signal and the Noise since I love chaos theory.  We are half way through summer already.  (crazy)

Enjoy!

Time to crack the books

Time to crack the books

Data Scientist or Data Analyst: What Do You Know?

Colette Grail:

I’ve done deliberate and contingency planning in the military using “things you know, things you don’t know, and things you don’t know you don’t know”. I’ve not seen it to describe data science versus analytics, but this is an excellent parable that explains the dark and the light and the grey in between.

Originally posted on Data Scientist Insights:

What you don t knowData sciences and data analytics not only use different techniques, that are often highly dependent on the distribution characteristics of the data, but also produce very different categorical types of insights. These insights range from a better understanding things you know you know (data analytics) to discoveries in area where you don’t know what you don’t know (data sciences). However, this knowledge metaphor can be a bit confusing, so I often use the “Darkness, A Flashlight, and the Data Scientist” parable. 

Flash Light

In your mind, picture a darkened room, where you are standing, but do not know where in the room you are. In your hand is large flashlight. You raise it slowly, pointing it in a direction. You turn it on and white light radiates forward.

The light of the flashlight shines brightly on a distant wall, where you see several items. These are the things you

View original 220 more words

Personal Health In The Digital Age

Colette Grail:

$300 BILLION price tag for non-adherence to prescription drugs? I’m not sure what that means. Who pays the price? Is it the consumer for a drug they didn’t take? Or the government or insurance?

Originally posted on TechCrunch:

[tc_contributor_byline slug=”brian-tilzer”]

We live in the digital age. You know that already. Two out of three Americans are now smartphone owners, and more than 86 percent of the population is connected online. But while digital has permeated everything from our social lives to how we work and how we shop, it is only starting to touch how we manage health.

Yes, nearly 70 percent of Internet users look up health information online (who hasn’t been on WebMD.com in a panicked moment of self-diagnosis?). However, only one in five of us have an app downloaded on our smartphones to track our health. And health apps comprised only 2.8 percent of total app downloads from the Apple App Store a few months ago.

All of this points to the disconnect between personal technology and personal healthcare, despite the vital importance of the intersection of the two. Personal technology is proliferating, yet the…

View original 915 more words

Fatigue Science Lets Pro Sports Teams Track Their Athletes’ Sleep

Tags

, ,

Colette Grail:

How well did you sleep last night? The NFL wants to know. Well, actually they want to know about their players. Sleep has been understood as part of a healthy routine for some time, but now quantifying the quality is a new grade of capability. Some NFL teams want to sleep to be a game changer.

Originally posted on TechCrunch:

As wearable activity trackers get increasingly smart and complex, Fatigue Science is measuring one thing and one thing only — how we sleep.

Fatigue Science’s Readiband looks very similar to a Fitbit or Nike Fuelband. It has a 3D accelerometer that tracks movement, impact, velocity, speed and frequency, a battery that lasts 60 days between charges, and it’s both water and pressure resistant.

The band alone is not a revolutionary development, considering that even the most basic wearable fitness trackers can monitor when you’re asleep.Screen Shot 2015-07-02 at 10.52.44 AM

Fatigue Science has the ability to detect sleep quality at 93 percent of the accuracy of a hospital sleep lab, but the real feat is their ability to predict human effectiveness and reaction time. The startup takes the sleep data captured by the band and runs it through a biomathematical model developed by the U.S. Military.

This level of accuracy may not be essential for most of us, but for elite athletes…

View original 380 more words

HIRE ME! In praise of “light quants” and “analytical translators”

Great article on the seam of quantitative analysis and decision making.

HIRE ME FOR YOUR ANALYTICAL TRANSLATION!!  (NO really, I’d be great.)

In praise of “light quants” and “analytical translators”.

In praise of “light quants” and “analytical translators”

Short Takes…on Analytics

A blog by Tom Davenport, independent senior advisor to Deloitte Analytics

It would be great if “heavy quants” also knew about business problems, and were fantastic tellers of analytical stories as well, but acquiring deep quantitative skills tends to force out other types of training and experience.

When we think about the types of people who make analytics and big data work, we typically think of highly quantitative or computational folks with hard knowledge and skills. You know the usual suspects: data scientists who can make Hadoop jump through hoops, statisticians who dream in SAS or R, data wizards who can extract two years of data from a medical device that normally dumps it after 20 minutes (a true request). Companies lust after these skills, and they are admittedly important and not easy to find. I’m singing the praises in this essay, however, of a different sort of analytical employee who is less widely sought by recruiters. When was the last time you saw a job posting for a “light quant” or an “analytical translator”? But almost every organization would be more successful with analytics and big data if it employed some of these folks.

The two jobs are related, but not identical. A “light quant” is someone who knows something about analytical and data management methods, and who also knows a lot about specific business problems. The value of the role comes, of course, from connecting the two. Of course it would be great if “heavy quants” also knew a lot about business problems and could apply their heavy quantitative skills to them, but acquiring deep quantitative skills tends to force out other types of training and experience.

The “analytical translator” may also have some light quant skills, but this person is also extremely skilled at communicating the results of quantitative analyses, both light and heavy. It would of course be great to have someone with heavy quant skills who is also a fantastic teller of analytical stories, but we are talking about a very small intersection of skills here. In fact, even if you once had strong communication skills, most graduate programs in quantitative fields will tend to drum those skills out. Academics communicate with each other in equations, stilted writing, and footnotes—none of which facilitate good storytelling.

These roles have traveled under the radar for some time, but they are beginning to be noticed. My friend Lori Bieda, who recently headed customer intelligence at SAS and now plays a similar role at a large bank, wrote a SAS paper on the topic.1 The International Institute for Analytics, of which I am a co-founder, argues that “the hot new job in analytics is storytelling.”2 This trend has not gone unnoticed at Deloitte, where the 2014 Analytics Summit was devoted to storytelling.

And some organizations and managers have embraced the light quant and translator roles. I recently spoke with Dr. Pamela Peele, the chief analytics officer of UPMC Health Plan. She’s a passionate advocate for clear communication of analytical results, and she did something about it, hiring someone to play the translator role for her organization. Peele has about 25 “heavy quants” in her organization, but she notes:

“When you ask a PhD statistician to write a report for the C-suite, it’s just not suited for their consumption. So several years ago I hired someone with a journalism background to improve the communication of our results. We had been generating great analytical studies, but they weren’t being translated into action—primarily because no one was consuming those results. The analysts still do the design and analysis, but it’s the storyteller’s job to get the analyst to tell them the story, and communicate the main points. The analysts will be on point 17 without any sense yet of how it will impact the organization. The storyteller’s job is to start at the end—begin with the impact, and then very selectively reveal how the result was achieved.”

Peele notes that some people have questioned the efficiency of having some people do the analysis and someone else do the reporting on it, but she says it’s been both efficient and effective:

“The PhDs are much more productive when they don’t have to spend time writing up their results in an easily understood form. And they’re much happier not having to spend time on it.”

I’ve met several other managers who believed in these light quant and translator roles and made them their analytical advisors. They felt that while these individuals would certainly need to consult with heavy quants on occasion, it was important to have someone who understands both business and analytics at their side. One senior executive at a large bank, who, at the time I interviewed her, was heading the distribution arm of the organization, commented:

“I have an MBA, but I’m not particularly strong in quantitative analytics. But for the last couple of years I have been pushing my direct reports to use analytical thinking in their decisions—for example, opening and closing branches, understanding customer wait times, understanding multi-channel customer interactions, and HR models for sales productivity, hiring, and attrition. In order to address these decisions, I needed to form much closer relationships with a team of quantitative analysts. Our bank has hundreds of quants, but until recently they rarely had close relationships with senior executives. I was one of the first business unit managers at the bank to establish a close working relationship with my analytics team. To do so, I rely heavily on analytical team leaders with moderate analytical skills, but who understand my unit’s business problems and can communicate them to hard-core quants. I have had good results with two such individuals, both of whom have Six Sigma and retail banking backgrounds. I seek quants who will push me to think a bit differently; I don’t want them to try to please, but to challenge my thinking and conclusions.”

This may seem an obvious thing to do, but it’s important to point out that most managers still don’t realize the value of light quants and translators, even though they may be frustrated in their dealings with heavy quants. Even at this executive’s bank, other managers argued (incorrectly, I think) that these positions were a waste of money and scarce openings relative to heavy quants and data scientists.

Organizations need people of all quantitative weights and skills. If you want to have analytics and big data used in decisions, actions, and products and services, you may well benefit from light quants and translators.

“In praise of “light quants” and “analytical translators” was published in DU Press on March 18, 2015.

New Process Can Print Stretchy Electronics Onto Your Clothes

Colette Grail:

An update on wearable technology.  Sensors that adhere to the skin and transmit their intelligence are not so far away …

Researchers at the University of Tokyo have created a single-step process to print conductive material on cloth, allowing manufacturers to build stretchable wearables that can test vital signs like heart rate and muscle contraction.

From the release:

Now, Professor Takao Someya’s research group at the University of Tokyo’s Graduate School of Engineering has developed an elastic conducting ink that is easily printed on textiles and patterned in a single printing step. This ink is comprised of silver flakes, organic solvent, fluorine rubber and fluorine surfactant. The ink exhibited high conductivity even when it was stretched to more than three times its original length, which marks the highest value reported for stretchable conductors that can be extended to more than two and a half times their original length.
Why is this important? Because it allows for the traces to and from electronic components to be amazingly stretchy. While components like chips and transistors are still hard to pull and bend, by allowing the connectors to bend and stretch in certain places you can create a tighter fit for measurement technologies and even bring connectors up close to your skin. The technology isn’t quite ready for prime time but it should be an interesting addition to the wearables world when it’s commercialized.

Originally posted on TechCrunch:

Researchers at the University of Tokyo have created a single-step process to print conductive material on cloth, allowing manufacturers to build stretchable wearables that can test vital signs like heart rate and muscle contraction.

From the release:

Now, Professor Takao Someya’s research group at the University of Tokyo’s Graduate School of Engineering has developed an elastic conducting ink that is easily printed on textiles and patterned in a single printing step. This ink is comprised of silver flakes, organic solvent, fluorine rubber and fluorine surfactant. The ink exhibited high conductivity even when it was stretched to more than three times its original length, which marks the highest value reported for stretchable conductors that can be extended to more than two and a half times their original length.

Why is this important? Because it allows for the traces to and from electronic components to be amazingly stretchy. While components like…

View original 61 more words

Using the Internet of Things to detect asset failures before they occur

Tags

, ,

This IBM post draws the connection between how the Internet of Things (IoT) is moving into individual lives with the capability deployed in big business to protect and maintain major assets.  In each case, sensors are utilized to detect conditions and alert to potential costly events.

Somewhere in between, small business can take advantage of the same principals – protecting equipment that is integral to operations and costly to repair.  If big business needs this to stay profitable, this is exponentially important to small business.

Using the Internet of Things to detect asset failures before they occur.

Follow

Get every new post delivered to your Inbox.

Join 132 other followers