Explore the deceptive nature of summary statistics in this 32-minute conference talk by Stefanie Molin, a software engineer and data scientist at Bloomberg. Delve into the world of data visualization and learn about Data Morph, an open-source package that challenges traditional assumptions by reshaping datasets while maintaining identical statistical values. Discover why relying solely on simple metrics like mean or standard deviation can be misleading, and gain insights into the importance of comprehensive data analysis. Follow along as Molin discusses the tool's development, its challenges, and potential future improvements. Gain valuable knowledge about data visualization techniques, shape class hierarchy, and dataset morphing processes. Ideal for data scientists, AI enthusiasts, and data engineers looking to enhance their understanding of data interpretation and visualization.
Data Morph: A Cautionary Tale of Summary Statistics
Overview
Syllabus
- Bio
- Introduction
- Visualization is an essential part of any data analysis
- Generate a target shape based on the dataset
- Shape Class Hierarchy
- Morph the dataset into the target shape
- Limitations and areas for future work
- Helpful resources
Taught by
Open Data Science
