UCSB Datathon/Data Day

A few months ago, I was searching for datathons to attend, where I landed on the the Southern California Consortium for Data Science (SCCDS) , leading to my discovery of the UCSB Datathon. So, Saturday, 3/7, I attended to UCSB Datathon to learn more about data science as someone presently studying it. Event Flyer

Panel 1

Jack Miller - Teaching prof. for stats and applied probability
Tisha Lwin - Data science program coordinator & advises undergrad/grad students
Daniela Trasladino - transfer from SBCC into UCSB
Nathalie Guebels - Computer Science professor at SBCC, has taught data science

What is data science?

  • Jack Miller - “It’s a beautiful marriage between statistics and the ability to code”
  • Nathalie Guebels - “It’s a place for multiple disciplines to come together and collaborate”
    • “data is not just numbers, its numbers about something” - domain expert (health for health expert, etc), it brings you to many different fields
  • Daniela Trasladino
    • the study of data to tell a story, make predictions Tisha Lwin
    • emerging field that can be applied in many different ways

What brought you to data science?
Jack Miller: returned to school and got a PhD to teach DS/stats
Tisha Lwin: graduated, chose math as a major, first year realized didn’t like math as much, went to econ and graduated in econ
Daniela Trasladino: fell in love with the class, wanted to see the programming in real world projects
Nathalie Guebels: started with electrical/computer engineering, went to a ds conference, and worked to incorporate it into the cs classes at sbcc, co-teaches

What makes a good data science program?
Nathalie Guebels

  • not name of school, but “are the upper division courses focused on things i want to do? capstone series? hands-on experience? projects?”
  • “Degree is just one line on a resume… you have to have it, but everything else on your resume is what is important to the job”
  • Internship programs, capstone series/projects, you can go into many areas
  • being happy while you do what you do is important Jack Miller
  • “you can go anywhere, and be awesome. It’s about how you fit in”

Tell us a bit about the transfer process.
Tisha Lwin: be proactive, take time to contact them and help setup course schedule
Daniela Trasladino: straightforward process, suggested study plan is helpful
Nathalie Guebels: keep track of coursework

What makes a student successful in your data science program?
Jack Miller: “If you want to be successful you will…Go to class, please. Stay on top of homework. Do homework right when its assigned, so it has time to perkelate.” + Office hours
Tisha Lwin: “For math and stats, students assume they can power through them. For any program really, it’s really important to work on your social skills when at university. It sounds lame, but forming a student group really helps a lot”
Daniela Trasladino: Take time to look for resources you’ll need post-transfer, find a mentor, or build a strong connection with a professor
Nathalie Guebels: Build your community, come to class to learn- the grade is just going to happen
Yan

  • don’t restrict yourself to learning just from the lecture notes
  • youtube, MIT, Harvard
  • combine them to really strengthen understanding of topics
  • campus has so many free services

Advice for building confidence?
Jack Miller: if you haven’t started something, that’s okay- start today. use your resources. “learning is a productive struggle.”
Tisha Lwin: You have more skills than you might think such as soft skills, other skills
Daniela Trasladino “Nobody was born knowing how to code, everyone can learn.”
Nathalie Guebels: Imposter syndrome is real, but “Everyone is here to learn.”

One piece of advice for students
Jack Miller: “Ask questions, enjoy the process, use your resouces”
Tisha Lwin: “Take care of yourself, youre getting your education for you, so you cant be in a class if youre struggling with other things”
Daniela Trasladino: “everyone goes at their own pace, and be patient with yourself”
Nathalie Guebels: “Don’t be afraid of trinyg something new, there’s no wrong major, don’t be afraid to reach out”

Lab 1

The first lab was learning R programmming. However, I already use R, and they used the same guide as the CSUCI datathon did. However, it was enjoyable getting to review (even if I use R weekly for my STATS C1000H class).
Presentation Slide

Panel 2: Careers in Data Science

James O’Brien - music composer, producer, director of a AI research lab (Sound Ethics)
Sam Shanny-Csik - Data science programs manager, UCSB bren school of environmental science and management
John Ramirez - Data scientist, works at MESA @ SBCC
Nancy Kim - Data analyst at AppFolio
Zoe Holzer - predictive data analyst of LA county department of public health

What is the most interesting part of your work
James O’Brien, Sam Shanny-Csik, John Ramirez: Working with students/recent graduates/undergrads/masters/phd
Nancy Kim: working on building things - agentic ai for example
Zoe Holzer: getting to solve problems

What does the typical work week look like?
James O’Brien: team meetings w research daily, requires a lot of collaboration, passion projects are the entryway to something new
Sam Shanny-Csik: lots of meetings, grading, lecture-prep
John Ramirez: discussion with his boss, plan the week, modeling data, creating spreadhseets every week
Nancy Kim: team meeting, a retro (debrief of the last week), 2-4 hours a week with stakeholders, rest of week spent with family
Zoe Holzer

  • data engineer/data scientist type work
  • similar to kim’s work
  • meet in mornings, full day to code, oftentimes meetings/respond to emails/plan
  • favorite part of day: coding, building product
  • works on linking records (to prevent duplicate records)

How important is domain expertise? Foundations/theory, or industry knowledge?
James O’ Brien: “One thing we can do is helping with domain experience for data science” Sam Shanny-Csik: “Bringing domain expertise is going to be more and more important” John Ramirez: Domain expertise is really valuable, data science has a lot of different backgrounds, a lot of final projects were “find your own data”, having expertise in another interest helped to do that Nancy Kim:

  • 5-6 years ago, she would’ve said technical skills are critical, NOW, her answer is domain expertise is hugely critical
  • AI is impact everyone’s lives, the coding and debugging is what becomes simplified, AI Agents are doing that now
  • training and building out semantic data models

Zoe Holzer:

  • math is still extremely important
  • math > domain knowledge, “side dishes of domain knowledge”
  • “AI cannot completely replace math and critical thinking”
  • the more you learn, the more you can abstract/model/picture things
  • some domain knowledge

If you do hire for data sci, what are you looking for a in a student? How do they stand out?
James O’ Brien: What they are interested in
Sam Shanny-Csik: build a portfolio of cool projects!
John Ramirez: teamwork, knowing how to collaborate with a team on a project, being passionate about your data, portfolio of projects
Nancy Kim:

  • soft skills
  • “are they somebody i want to work with/is it someone i would trust to work with the data + how is their curiosity to learn”
  • curious + drive to learn
  • if you can incorporate AI into portfolio, even very simple things

Zoe Holzer:

  • being able to communicate interest + passion
  • deeply understand what you’ve done
  • for technology, really important is resource management, big data computation, and parallel processing
  • optimization, help things run faster

One piece of advice?
James O’ Brien: Keep learning, stay motivated.
Sam Shanny-Csik: Join in-person learning communities! + network
John Ramirez: Build a bigger network, don’t be afraid to fail, don’t ever give up.
Nancy Kim: Own your mistakes, don’t be afraid to fail.
Zoe Holzer: “a lot of you won’t like this answer… but you really need to study.”

Lab 2

The second lab was the datathon, where we got 1.5 hours to code and make our analysis. Everything I made is available in this repository, but essentially, I answered some of the challenge questions and the 1/2 main questions. While the R programming wasn’t anything inherently new, I think practicing quick thinking and critical analysis was definitely worth coming to this event. Plus, there was some R syntax I had to look up, like ~facetwrap, or na.rm. I enjoyed talking with the various professors and data science students who came and observed the event as well, they were incredibly helpful and informative as to what data science is like at UCSB.

Overall

After the share-out where I presented my findings to the group, I enjoyed the long ride home along the coastline. All in all, I feel I practiced and learned new things, especially from the panelists’ insight. Thank you so much to the incredible team who put on this event!