Andy Pulkstenis is a Director of Advanced Analytics for State Farm Insurance in Bloomington, IL. In this role he leads a team of advanced analytics professionals providing data science, statistical analysis, and predictive modeling support for the enterprise across a variety of business units.
Andy received his MS degree in Statistics from Penn State University and holds a BS in Mathematics (Statistics minor) from Messiah College. He has spent over 20 years in applied data science with Trilogy Consulting, Capital One, and now State Farm, with particular focus on predictive modeling, experimental design, building & developing analytic teams, and statistical recruiting. He serves on advisory boards for Corinium's Chief Analytics Officer Forum and the DMA's Marketing Analytics Conference, and recently received the Casualty Actuarial Society’s “Certified Specialist in Predictive Analytics” designation. When away from work, he enjoys spending time with his family and NHL hockey.
As the practice of applied Machine Learning grows and data science/advanced analytics business teams proliferate, there have been correlated increases in operator error and questionable practices. This presentation details some common pitfalls and myths to watch out for as you grow or build your data science divisions and presents compelling facts to counter growing misperceptions as more and more businesses try to leverage these techniques. Grouped into conceptual, analytic, and organizational pitfalls and outlining core concepts critical to understanding machine learning as a powerful business tool, this engaging presentation shares examples and lessons culled from a combination of experience, published articles, and the broader data science and advanced analytic communities.