Data Science in the Wild Wild Midwest

Rooms 301 and 302 August 22, 2016 10:05 am - 10:35 am

Bookmark and Share

Bo Peng, Datascope Analytics

Bo Peng, Datascope Analytics

Download Presentation

As a data science consultant, I use methods from statistics, machine learning, and design to solve business problems. But business problems rarely come in well­-defined, prepared packages. So while we could theoretically spend ample time fine­tuning our models or algorithms, it is far more worthwhile to spend our efforts communicating with the clients to make sure we solve the right problem in the first place, and continue to pivot if need be. In this talk, I’ll go through how we do this in practice, with a few examples of projects: mining unstructured text and creating topic models to figure out what people are good at; automated classification of legal documents to determine what’s relevant to a case or not; designing and teaching an end­-to­-end bootcamp for smart, quick learners to become entry­ level data scientists in twelve weeks.