An Evening of Lightning Talks with pydata
Took some notes at the Pydata meetup tonight.
Schaun Wheeler from Valassis Digitial
How Design Skills are necessary in Data Science
Good Implementation Requires Good Design
- Know it when I see it vs here are the rules you must follow them
Having vision in your design
- Know your tools
- Knowing how it all goes together
Many of the technical aspects are being automated
Most fundamental tools are now part of an api Design is taking the end goal
- breaking down
- difficulty is in the conceptual experience
- Domain Expertise is the hardest part to learning
- Communication is required
- Design is a spectrum
Anthropologist gather stories and interpret them in a way that is understandable to the reader.
- Iterative process
- Work through the variations
- Each round the difference is smaller and smaller
- Each step is a difference in how the designer sees the world vs how others see the World
The only way data has a voice is if the tool is built to enable that.
- Tooling for design is an unexplored field.
- Developing data analysis soft skills will require tools that leverage those skills.
A focus on design is the best chance that data science has to survive scrutiny from a skeptical public.
Triangle Computer Vision Image Recognition Meetup
Discuss Cutting Edge Technology
- Last speaker was from iRobot
- Learning in Computer Vision and Image Processing
- Currently going over a Udacity course
Triangle Machine Learning Day
“All explanations are wrong, but some are useful.”
Nick Haynes (2018)