Data Science Expertise Comes in Many Shapes and Forms

Few things bring us more joy than publishing an author’s first TDS post. Part of it is the thrill of discovering a new voice sharing their unique experiences and knowledge. Another is realizing, time and time again, that there’s always a fresh angle to explore, even when we deal with common questions and challenges.

We’re dedicating this edition of the Variable to recent contributions by our newest cohort of writers, and we’re doing it for a couple of reasons. First, we wanted to celebrate their ability to communicate complex ideas with clarity, precision, and a distinct personal perspective.

Second—and perhaps more selfishly—we’re highlighting these articles in the hope that other members of our community feel similarly inspired to share their expertise with TDS’s audience. We cover a very wide range of topics—from AI ethics and climate-change data to ML pipelines and product management. All that really matters is that you care deeply about your subject matter and think others should care about it, too.

Happy reading!

• "Does the world need another introduction to the GPT family?" is a fair question to ask. We’re pretty certain your answer will be a resounding "yes!" after reading Beatriz Stollnitz‘s thorough explainer on the core concepts behind these ubiquitous language models.
• Relational databases have been around for decades, but their longevity is no reason to dismiss their power and time-tested reliability—on the contrary! Stephanie Lo explains why data scientists would be wise to develop a deeper understanding of how databases work and learn how to use them effectively.
• As a practicing NLP researcher, Francisco Caio Lima Paiva was curious to see how well ChatGPT might perform on sentiment-analysis tasks, so he set out to test the chatbot’s abilities against those of domain-specific models.

• Do you occasionally feel intimidated by the idea of writing custom functions? Don’t miss Vivian Peng‘s new post, then: it covers the full process of generating multiple plots in R using a custom function. As an added bonus, the main steps Vivian describes can be generalizable across a diverse range of workflows.
• "Working with data that has dates and times can be easily overwhelming," says Andreas Lukita; we suspect many of you are nodding your heads vigorously at that statement. Andreas is here to help, with a comprehensive, one-stop resource on Python’s datetime module.
• Managing ML and AI teams comes with its own set of rewards—and challenges. Przemek Pospieszny shares insights based on his long career across both R&D and product leadership, and stresses the importance of agility, planning, and finding the right balance between innovation and pragmatism.

Thank you for supporting our authors! If you enjoy the articles you read on TDS, consider becoming a Medium member – it unlocks our entire archive (and every other post on Medium, too).

Until the next Variable,

TDS Editors

Share This Article

• Share on Facebook
• Share on LinkedIn
• Share on X