All Hands on Data #31
Thank you for joining us on our voyage through data this year! We look forward to continue navigating the sea of articles with you next year.
Why You Need To Be Using Fake Data
An interesting thing worth noting for those working in data-heavy fields or roles is that you need to test your pipelines while also adhering to data integrity and compliance demands. Using fake data to ensure your system is functioning properly is a great, although rarely used, practice, according to the author. - John Forstmeier
The Data Compass: How to Build a Data Strategy
Dan proposes a model for creating an organization data strategy, the "Data Compass", where you data can be pulled in any direction. In your organization, data can be seen as a Culture, a Differentiator, a Capability, or an Asset. These forces push and pull on each other, but each can be right in their own regard. You just have to evaluate the merits of each and decide what makes sense for your business. - Blake Burch
FTX Implosion Highlights the Importance of Conversational AI
I've always been a bit skeptical of crypto - and the recent fall of FTX, or should I say implosion, has led to to dig in and learn a bit more. One of things I hadn't considered is how these companies are managing internally with things always going up and down. I wonder what level of customer service you'd receive with questions or concerns. Maybe moving to AI for chat bot like tools is the answer? - Joseph McDermott
Solving Multi-Armed Bandit Problems
I would love to see a raccoon loose in a casino and all the mischief they could get into. The numerical value of regret is especially interesting because it is usually difficult to quantify regret in real world decisions yet is very obvious in mathematical examples. - Eric Elsken
Open Source SkyPilot Targets Cloud Cost Optimization for ML and Data Science
SkyPilot is an interesting open source tool that could help users save money. The idea of a tool that automatically helps find the cheapest availability zone, region, and provider for the requested resources is promising. - Jon Davidson
2022 Data Science Research Round Up: Highlighting ML, AI/DL, & NLP
Gutierrez provides a great compilation of papers about machine learning, AI, and nature language processing that were published over the past year. Catch up on what you missed and definitely read my favorite "TalkToModel"! The authors aim to use NLP, packaged in an interactive dialogue interface, to help explain ML models. - Katt Baum