Comments
SAS's Schabenberger Shows How the Pieces Fit Together
View Comments: Newest First | Oldest First | Threaded View
Page 1 / 2   >   >>
Re: pieces
  • 12/28/2016 12:34:24 PM
NO RATINGS

@kq4ym, have to agree with your take on projecting into the future. The old educated guess is practiced under many other titles but no doubt is the pervasive MO.

Re: pieces
  • 12/28/2016 9:43:43 AM
NO RATINGS

It's always a good problem to look at; just what happened in the past that was really useful ...and will lead to future successes? Without that elusive chrystal ball, it's pretty hard to really know of course. Things we thought were important might turn out to be not so, or even false. The most we can do is probably take some educated guesses at the probability of what we learned by able to be propelled to greater knowlege in the future.

Re: pieces
  • 12/26/2016 1:34:37 PM
NO RATINGS

The issue is complex because the past doesn't always predict the future hence, using past behavior without verifying change becomes troublesome. The modeling needs to account for evolution and change.

Re: pieces
  • 12/26/2016 12:42:08 PM
NO RATINGS

@Loisw900, I'm not sure that I can go as far as saying major overhaul, but more scrutiny and continually improving the process is how I would view and state it.

Re: pieces
  • 12/26/2016 12:37:37 PM
NO RATINGS

@impactnow, I also thought about the miscues that played out on the election and the impact as it was on such large stage. Short term embarrassment, long term gain because it started a conversation which will have analytics prevail with a deeper understanding by a wider audience.

Re: pieces
  • 12/26/2016 9:25:51 AM
NO RATINGS

And like most refrigerators in homes, we look past those pieces focused on what's behind that door.  We're realizing that we must take a step back to see what those pieces are suppose to be.

Re: pieces
  • 12/23/2016 5:39:33 PM
NO RATINGS

@impactnow   Excellent point.  Modeling and Predictions both need a major overhaul in both thought and practice. I am not sure this is a battle that will ever be completely won, since variables change by the minute.   But if we are to use Models and use them to make decisions then the constraints for them need to be thoughtfully reconsidered.

 

Accepting "model bias" will only get you so far.  

Re: pieces
  • 12/23/2016 5:33:57 PM
NO RATINGS

This Big Picture approach going forward seems natural to me.  There are many tools to use but it is the collection of results from these tools that needs to be understood and place correctly together to give companies  a "big picture".  

This should be one of the main focuses for 2017 and beyond.    Though I am still not convinced the battle to convince businesses to use and invest in Analytics is over and it probably never will be.

Re: pieces
  • 12/23/2016 5:12:44 PM
NO RATINGS

It really was an interesting year for analytics, we were sobered by many analytics mishaps including the election predictions. We are still trying to get our arms around all the possibilities of modeling and predictions and still see that they are not perfect. In 2017 I hope that we will use lessons learned to better target our modeling and predictions to create more value and less disappointment. Then the pieces will fit together and also make sense!

Re: pieces
  • 12/22/2016 1:41:26 PM
NO RATINGS

@Jim an interesting way to look at it: a picture of things that don't fit together like pieces of a puzzle but that do add up to a whole. 

Page 1 / 2   >   >>


INFORMATION RESOURCES
ANALYTICS IN ACTION
CARTERTOONS
VIEW ALL +
QUICK POLL
VIEW ALL +