This is a great video and the panel really describes the role of the data scientist. They talk about how it gets used and misused at times in the real world like running parametric queries on non parametric data for one example, and people doing really “silly” averages of averages the one panel member states and she works in data at T-Mobile, but it’s all over the place and nobody has enough data scientists that can visualize. I agree with that as before we had “big data” around 10 years ago or more I was visualizing, pretty much in my head and thoughts as we didn’t have the tools we have today that puts this in front of us as a “no brainer”.
Joining non summarized data is another fallacy that comes up and you see that all over the place as folks run a query to see what sticks to the wall and then sell it and sometimes there is no value if it doesn’t do anything, or in the case of how we seem to see everything today, money. The panels goes beyond this this though, thank goodness and talks about the “real” value and not just some algorithms created to make money. We saw that and I wrote it up a couple of times with the FICO analytics they sell. It has no value as credit scores and data mined from the web (non credible) has no value when you bring this down to an individual level. Sure there’s no problem using it to predict in numbers, but again what they did was unbelievable and will deny people medical care. Why? The used non credible data and on social networks anyone can lie their fanny off and yet they are out there selling this ridiculous analytics service to insurance companies and to pharmaceutical companies to predict of you will be a patient that will take your prescriptions. I give that gal on the panel a lot of credit here to be bold enough and tell it like it is. The FICO example is where Algo Duping begins and I do hope that when it comes to predictive analytics that insurers and pharma companies listen to somebody else when it comes to analytics that offer value and not this.
FICO Analytics Press Release Marketing Credit Scoring Algorithms to Predict Medication Adherence–Update (Opinion)
In addition on the panel is a woman from Archimedes who does healthcare modeling. Archimedes was started at Kaiser Permanente years ago and then branched off at their own company. Kaiser is still a client/partner and we have seen some similar data work coming from within the organization as well. The first part of the video drags so to get to the meat and potatoes quickly advance to around the 35 minute mark. You need to investigate when a data scientist, aka advanced analytics is needed when the ultimate business intelligence goal is not defined. Most companies run traditional business intelligence reports including most of the big guys. The video actually mentions Quants who “are” the data scientists.
The gentleman from NASA has a lot to add as well when he speaks of the average scientist being able to use what they build. He says you need a lot of citizen scientists and surprisingly he states in what they do there’s not a real heavy use of quantitative data. Business analysts are being trained in college but people are learning on their own through visualization tools. I learned a new term here “Algorithm Jockey” and they are out there. There’s a nice question and answer at the end that talks about the time required to “prepare” data, and yes that is needed and this goes back to my coined phrase “the short order code kitchen burned down a few years ago and most missed the fire sale”.
Now to finish here, here’s 4 videos that will tell you what Algo Duping is all about, where corporate America has both by plan and some by accident created analytics and a lot done with some big data resources have used algorithms to move the money to the one side where the wealthy have it all.
Where do you start and where do you pitch is also discussed. Get that executive that is not in a cave or sitting in the 70s and get them to move with it. Now one last thought and here’s a link back to a post from the Medical Quack over 3 years ago, August of 2009…..what do you think? That post below is and was visualization at work and you start thinking this way when you know the mechanics and can visualize.
As this moves quickly should someone be minding the shop? We had Mary Shapiro state she was leaving so a person that is part “computer scientist” running the SEC in the future would certainly be nice and do more for the financial status of this country than what has been done in years. Congress is a big part of that issue too with a huge plague of digital illiteracy and drama queen antics that seem to get the best of them to where they can’t think about visualizing. BD
Does big data live up to the hype? Yes. To me, big data means technology and business alignment---that Holy Grail endlessly pursued by CIOs---becomes a no brainer. Big data projects by nature are about revenue, risk and profits. In other words, IT and the business can't help but be aligned.
Clearly, we're in a big data hype cycle that I put on par with the Linux and open source software craze in the late 1990s and early 2000s.
Back then, Linux was going to change the world, kill Microsoft and other things. In many respects, Linux and open source software (Android for instance) did change everything. But a funny thing happened on the way to revolution---open source software became commonplace in every data center and now is take for granted. The revolution happened, but we just stopped talking about it as much. Cloud computing is playing out in a similar fashion.