"Dan Barbatti's Data Science/Machine Learning Page"

Dan Barbatti and his boy Rufus





Welcome to my site where I hope to discuss my journey in Data Science to help others along the way !

A little background:

My name is Daniel Barbatti and I live in St. Louis, Missouri. I Started in Information Technology as a vocation around 1978 working for the University of Missouri in the Urban Information Center.   At the Urban Information Center our speciality was processing demographic information especially Census data.   So my first I.T. Job was "Data Science" mostly on the "descriptive side".   I enjoyed working there, especially with the academic staff for a few years until leaving for a job position at Southwestern Bell Telephone (then a part of AT&T).  The job there was in the "Capacity Planning" group performing forecasting using mainframe utilization data.   This was my first experience into predictive models and I found it very rewarding.   Although now working for a contracting company I am still at AT&T after having many different positions and of course always dealing with data.     

Several years back I started down the road to learn to learn more advanced modeling.   In the past I had just used various linear regression techniques for the predictive work.   With the recent growth in hardware and software capability there are a lot more methods available and the intent of this page is to share the journey.              

The Journey

In college I had taken a course in Artificial Intellegence where a group of us created a program to play checkers in LISP versus the mainframe. Thankfully the machine learning that currently thought of as "AI" is a lot easier! Many are afraid to make the jump because of the math and stats background they think they will need. Obviously the more math and stats you have the more you will understand what is going on "under the hood". That can be an advantage but not having an advanced math degree is certainly not a show stopper! In my opinion an ability to understand the business problems you are trying to solve and an ability to commmunicate to the client is more important. These days there APIs available that make it straight forward to create machine learning models without knowing the nuts and bolds. I once heard Cassie Kozyrkov mention if are running a restaurant you don't need somebody that knows how to build a microwave, you need someone who knows how to use one!

Below is the first video in a wonderful ML intro Cassie does on her channel: 

 

Some Resource Recommendation

 

Dan Barbatti's Web Profiles

Personal Web Sites (self developed):

 

Other Links About Dan Barbatti

My water garden work:

My work in Information Technology:

  • MXG Chapter 99 - CODE SHARKS
    A site where I was given credit for coding enhancements I provided to the MXG product. It was a significant achievement to be a contributor.

My certication and study of Tai Chi :

My work for charity: