Statistics for Data Science: Basics to Production

Updated Title: Using ML & Applied Stats to Tell a Data-driven Story

This workshop focuses on using applied statistical learning to produce a data-driven story. Using a previously cleaned, analytic-ready dataset with practical machine learning tools, we will interpret and discover fresh insights to tell its underlying story. Example case (tentatively) is recent prisoner recidivism data for Broward County, Florida from recent ProPublica investigative reporting; original datasets available on Kaggle. Tools explored range from clever visualizations for probability progressions, to Shapley plots, then to cloud Google TF What-If Tool and AWS Clarify explanations.

About the speaker:
Mash Zahid is a management consultant who has led development of artificial intelligence systems that leverage state-of-the-art infrastructure and machine learning methods for financial services, healthcare/biotech and retail industries. Mash was a global strategy director for AI at KPMG, and also served as Associate Partner leading the Machine Learning Practice for a Microsoft services provider. Previously, he was a senior strategy manager at The Home Depot, where his data-driven approach for improving retail operations and People strategy was detailed in a Harvard Business Review cover story.

Mash has an MBA in Analytic Finance and Accounting, with PhD coursework in Behavioral Finance, from The University of Chicago Booth School of Business. He received a BA from the College of Wooster with his senior thesis on “Political Participation in Cyberspace."