The vertical set of tabs summarizes some of my practitioner and observer knowledge of Predictive analytics and related areas. Let me say this up front: there is no one best algorithm or approach. In some cases one cannot even define best. In other words, There is no free lunch!

What you'd find is a reasoned and pragmatic discussion of many popular techniques and technologies, and resources to learn more should you so desire. I am not an academic, so the pedagogical urges compete with my work life and family and might often lose.

Below on this page you'll find overt promotion of my personal projects that could be of interest to some. Please don't send me money.

To the left a wordle tag cloud of the book I wrote on search monetization in my spare time in 2012. The book won't be published; however, the visualization is fascinating and captures the essential problems in online advertising and web user behavior well.

Maybe it was the Engineering ethos, but for some reason I never used S-Plus and until two years ago, R. As I started learning R, my objective was simple. I knew what I wanted to do, having done it in MATLAB or with shell scripts or with SAS or JMP. I just did not know how exactly to do that in R. Once I got some familiarity, and noticed how R allowed X11 colors, I was hooked on prettying up the figures to my liking.

I took some notes that have since grown to be about 80 pages in LaTeX and I'm working on creating an index. These notes are something I plan on making available on my website in Fall 2013, at no cost to those wishing to learn how to do Statistical Modeling in R. To make it complete, I've added some math to make some points more precisely.

Being task-focused but not a cookbook, the first set of chapters I am close to finishing demonstrates how to use the same data set to build linear regression models with traditional (frequentist), Bayesian and M-Estimation (Robust) methodologies along with Regression trees.

Classification of course will be a harder problem to address due to additional considerations, but it'll be fun.

From 1998 to 2003, I researched Nonlinear and Chaotic Dynamics (informally known as Chaos theory), while a graduate Research Assistant at University of Tennessee, Knoxville and Oak Ridge National Lab. My doctoral studies there were funded by MCEC, a NSF Industry/University Cooperative Research Center (I/UCRC) .