Session: Fake Data, Real Results: Using Synthetic Data When You Can't Use the Real Stuff
ARE YOU...?
• Wanting to do more with AI than use ChatGPT to write Linkedln posts and respond to emails you’re too busy to read
• Fed up with preventable bugs getting to pre-prod, caught too late and by your users, frustrating internal customers and casting doubt on SQA quality
• Pretty sure synthetic data and dummy data are the same thing (they are not, by the way, and I can prove it with colorful charts and math!)
This is a great talk for those who are data-curious, but do not (yet!) consider themselves data scientists or analysts. In this presentation, I take you through a data science research project for loans. See how I solved several real-world financial industry business problems with synthetic data. We will make true data and synthetic data battle it out in real time to see how they perform in predictive models. Then, you can take a peek under the hood to inspect the code I used so you can replicate a similar project on your own.
Bio
My family describes my job to friends as "spreadsheets and bots," which, to be fair, is not that far off the mark. I am a Certified Scrum Master and Product Owner, and recovering web developer with years of experience collaborating with distributed international development teams across multiple time zones. I currently volunteer on the Steering Committee for Navy Federal Credit Union's Scripting Community of Practice. In addition, I am a former Workshops Director and Conference Organizer for the Code(Her) Conference in DC, "Millennial Women in Tech" roundtable panelist, and legacy leader for the former Women Who Code non-profit. You can call me Ann or "Nova" because my legal name is 24 Scrabble points.