Thinking Analytically

A Beautiful WWW posted a great primer to machine learning. Among the recommendations:

• The Elements of Statistical Learning: Data Mining, Inference, and Prediction, Second Edition (Amazon or free PDF)
• The open source stats package R
• Some data sets from UCI Machine Learning Repository
• The MIT OpenCourseware machine learning course

Though that’s enough to get you started, the author promises to add on as time goes by.

As a Freakonomics fan, I can’t help but be intrigued by the “freakquel” available for pre-order on Amazon.

SuperFreakonomics challenges the way we think all over again, exploring the hidden side of everything with such questions as:

• How is a street prostitute like a department-store Santa?
• Why are doctors so bad at washing their hands?
• How much good do car seats do?
• What’s the best way to catch a terrorist?
• Did TV cause a rise in crime?
• What do hurricanes, heart attacks, and highway deaths have in common?
• Are people hard-wired for altruism or selfishness?
• Can eating kangaroo save the planet?
• Which adds more value: a pimp or a Realtor?

I’ve ordered my copy. You?

Analytical Thinkers typically are searching for the truth and the truth often implies hard, cold facts. The NY Times article Mining the Web for Feelings, Not Facts might seem hypocritical but speaking from experience, “there’s gold in that ther’ text.” In the mid ’80s I managed a group of employee relations analysts that mined mountains of text in an attempt to quantify employee morale for a 10,000 person company.  Out of this work came “ERATS” or the Employee Relations Attitude Tracking System.  (We were really tracking morale but couldn’t come up with a good acronym with that pesky “M”. Sentiment analysis is exactly what we were trying to do but completely by hand.

An emerging field known as sentiment analysis is taking shape around one of the computer world’s unexplored frontiers: translating the vagaries of human emotion into hard data.

Two recently devoured books on my shelf explore the web-analytics from a few angles.  If you haven’t already, check out:

Click: What Millions of People Are Doing Online and Why it Matters

I’m a bit behind sharing my recent reading list so I’ll try to spend some time over the next few weeks providing some recommendations.

I “listened” to the audiobook of Click: What Millions of People Are Doing Online and Why it Matters on my iPod.  This is one of those fasicinating books that makes one think, “How can I get a job like that?”

Bill Tancer mines the gold found in search-engine data.  I imagine him sitting in front of a computer screen with a massive amount of data starting his day by saying, “OK. What can we learn today?” What he learns is fascinating.  From prom dresses to porn to politics, he uses search data to understand and predict consumer behavior.  No need to search for this book — just “Click” on the image and get a copy for yourself!

Though not an analytics book per se, Tribal Leadership: Leveraging Natural Groups to Build a Thriving Organization contains some interesting analysis on the types of groups found within organizations.  I found it an interesting and valuable read.  Actually, I didn’t actually “read” it.  As my wife says, I cheated by listening to the audiobook.  For a limited time, you can download the audiobook for free from the Zappo’s web-site.

I just started reading Outliers: The Story of Success on my iPod touch. It’s the first Kindle format book that I’ve purchased.  So far, I like the experience.  Here’s a riddler from the book:

Why do most Canadian hockey stars have birthdays in January, February or March?  Probably has something to do with being born in the dead of winter and better able to tolerate the cold and icy conditions.  Right?

Would that explain why Czechoslovakian soccer stars have birthdays in January, February or March?  Then why do almost twice as many US major league baseball players have August birthdays than July?  Why don’t birthday patterns show up for football or basketball players?

Hold that thought.

Gladwell goes on to demonstrate that it typically takes 10,000 hours to truly master a professional level position such as pro athlete, musician, ballerina, etc.

An athlete born six months after a future star starts only 4,380 hours in the hole.  Over time that would seem irrelevant.  However, the sports mentioned above typically have a age cutoff date for young teams.  Soccer and hockey cuttoffs are typically January 1 so to players born only a day apart (say Dec 31, 2001 and Jan 1, 2002) will be on two different teams.   And that “younger” one will be the oldest on the team so after a few years he or she could be significantly larger and more mature that the other teams.  Often, these individuals will get special attention including spots on all star or club level teams.  Over time, they’ll put in significantly more time than the players born at other times of the year and will get to the magic 10,000 hours sooner.

Makes one wonder about the validity of cut-off dates.  Any ideas for a better system?  Do similar practices have implications outside of the sporting world?

I was first on the list at the local library for the new book The Numerati so I decided to take it on a business trip last week. I started reading on my first flight and by the time I reached my layover airport, I decided to buy the book at an airport bookstore. Of course, I paid full price but at least I could start writing in the margins!

I’ve started a mind map of some of the key concepts. It is interactive in the frame above or click on the map to collaborate and add your own information at Mind Meister. Just create a free account. Or, for less than $50 you can create a premium account which will allow you to create an unlimited amount of shared maps. This seems like a neat opportunity to try a relatively new form of collaboration. Tell me what you think in the comments.

Link to Amazon.com

With NFL football season starting up this weekend (way to go Packers and Aaron Rodgers!), it’s hard not to think about all that cash floating around on the field trying to avoid injury. I don’t know the details of Tom Brady’s unfortunate injury but the it’s hard not to think about the danger QBs face when a few hundred pounds of meat eating muscle is rushing towards them — especially from the blind side.

I enjoyed reading The Blind Side: Evolution of a Game by Michael Lewis last year after I read Moneyball. Though the story is about a particular athelete, Lewis starts by discussing the importance of the left tackle position and why it is one of the highest paid in the NFL.

Here’s some interesting arm chair analysis to support that claim.

A few months ago I mentioned that the bookHow Doctors Think reminded me of an article I had read discussing analytics involved in bird identification and disease diagnosis.  While going through some old files the other day in preparation for an office move, I actually found the original article!  It wasn’t published in a birding or nature magazine.  It wasn’t published in a medical journal.  No, it was published in the August 2002 issue of Harvard Business Review!

The article is a Q&A with birders David Sibley and Julia Yoshida.  Sibley is one of the foremost birding experts in the world.  (I met him a few years ago and own several of his books.)  Yoshida is both a birder and medical doctor.  In the article they discuss the importance of pattern recognition to their respective fields and HBR makes the connections to business.

I found a number of insightful concepts in the article — too many to mention in one post.  Over the next few weeks, I’ll share a few of them.  Here’s a starter:

The ability to grasp complicated phenomena and discern possible trends from seemingly random events can be a source of competitive advantage, allowing managers to capitalize on opportunities before they are apparent to others…To thrive in an information-rich world, executives need to be adept at pattern recognition.  They can learn a lot from expert-birdwatchers.

You can purchase the article at the HBR web-site.

Amazon carries the books mentioned in the article including:

I read How Doctors Think this weekend while working in the yard.  (OK, I was listening to the audiobook on my iPod.)  Fascinating book!  The book is based on the personal experience and field research of the author Dr. Jerome Goopman.  It delves into the decision making process — sometimes faulty — of the modern medical doctor.  To me the book was one part fascinating and one part scary!

It made me think of a Harvard Business Review article I read several years ago that discussed the analytics of bird watching!  (I’m an avid birder.)  The connection?  One of the interviewees was a doctor and the article drew parallels between analytics in business, identification of birds in the field and diagnosis of illness.  I’ll need to find that one and share it.