In this session, we will examine the opportunities for decision making while building a data analysis pipeline and follow the consequences of those decisions for the interpretability of the results. In addition, we will dive into examples of various types of bias, as well as examples of assumptions made in data collection, in implementation and in statistical modeling.
Throughout the session we will be discussing what to consider when choosing your quality control measures to maximize the trust that you can put in your data. In order to keep this session interactive, we will design a mini analysis pipeline using open data from Miami-Dade County. The title of this session is intended tongue-in-cheek since many of the key decisions that go into building a data analysis pipeline are context-dependent.
Speaker Bio: Dr. Hadjixenofontos joined the University of Miami Center for Computational Science in 2016. As the Center’s Director of Engagement, she leads a number of programs that aim to support the development of computational skills and adoption of computational mindsets in various populations. She’s particularly excited by the science part of data science, as it relates to assumptions, bias and their relationship with asking questions that make sense. She holds a PhD in computational genetics from the University of Miami John P. Hussman Institute for Human Genomics.
Meetup Info: This Meetup is a friendly environment where all questions are appreciated and where members learn from each other.
All experience levels are welcomed. Join us! | Bring a laptop to maximize participation.
Special thanks to StartHub Miami (http://www.starthubcenters.com/) for being the location sponsor!
Vehicles entering the designated parking lot after 5 PM, pay only $3 for the rest of the evening. The lot is located immediately east of 66 West Flagler, where we’ll meet.