Data Community DC, Inc.

For those in the DMV, Prema Roman will be presenting at Data Community DC, Inc.'s Full Stack Data Science event on the evening of July 23. Don't miss it!

Join the Full Stack Data Science Meet Up May 30th at 5:30 p.m. ET to discuss AI Model Orchestration with the Mage Framework and building Generative AI Bots to play video games! Get the details and get in on the fun: https://lnkd.in/e8DXKGX7

DC data and API folks: I'll be speaking virtually to the DC Data Wranglers on "Developing APIs that Data Scientists Will Love" May 20 over lunch. Hope you can join and pick up some tips from research for my O'Reilly book Hands-On APIs for AI and Data Science that will be published May '25. If you've wanted to see FastAPI in action, you'll enjoy this one. #datascience #apis

I'm excited to be presenting at the 10th annual New York R Conference in 2 weeks! (May 16-17) Thank you to the conference organizers for giving me an opportunity to speak about "Ragtime in the Big Apple: Chat with a Decade of NYR Talks," where I will demo an app allowing you to ask questions of the past 10 years of conference talks, and give an intro to Retrieval-Augmented Generation (RAG) - a popular method allowing users to integrate Large Language Models with their own documents. I'm honored to be speaking alongside data science leaders as Abigail Haddad, Hadley Wickham, Jared Lander, Andrew Gelman, and many others. Hope to see you there! If you can't make it to NYC but still want to participate, they have a virtual option. Don't forget to take advantage of the special 20% discount on your ticket by using the promo code NYR10 at https://rstats.ai/nyr.html. (In case you heard me talk about this a few months ago at the Bethesda Data Science Meetup, I've got a few neat improvements, so you might want to check it out again 😉 ) Let me know if you plan to be there! #rstatsnyc #rstats

Are we still training machine learning models on cats versus dogs? While clichéd datasets like iris flowers, Titanic survival records, and MNIST handwriting samples have paved the way in machine learning, they often fall short in demonstrating the real-world utility of ML technologies. This gap poses a significant hurdle when stakeholders try to visualize how these basic models could solve specific business challenges. Utilizing familiar datasets such as cats versus dogs may streamline initial machine learning projects but can obscure the true potential of your solutions. Presenting these models to stakeholders often requires them to make a considerable leap from straightforward examples to complex, industry-specific applications. This not only challenges their understanding but also risks diminishing their interest, as they struggle to see the practical relevance of the technology to their own contexts. To ensure our machine learning models reflect real-world complexities, consider these alternative approaches to dataset selection: 1. Industry-Specific Datasets: Seek out or custom-create datasets that address the specific challenges and nuances of the target industry, ensuring the training data closely mimics real-world scenarios. 2. Synthetic Data Generation: Use synthetic data when real data is scarce or sensitive, creating diverse and extensive datasets that respect privacy concerns and are tailored to specific needs. For instance, instead of using MNIST for handwriting analysis, consider a dataset derived from real-world documents in legal or medical fields, where the variety and stakes are much higher. Alternatively, replace the simplistic iris dataset with complex image datasets from agricultural fields for crop disease detection—these provide greater relevance and challenge, aligning more closely with end-user needs. Embracing sophisticated datasets that mirror real-world complexities is key for showcasing the true utility of machine learning models. By moving beyond traditional, simplistic datasets, we bridge the gap between theoretical applications and practical solutions. What datasets have you found most effective in your projects? Share your experiences below! #AI #MachineLearning #DataScience #Innovation #TechTrends

Great way to start the week! Not only was there the #eclipse, but Vivi Poit visited the Arlington Tech STEM girls at Arlington Public Schools. She shared her software engineering journey, gave encouragement & answered questions. Thanks to Michelle Van Lare & Monica Lozano Caldera for hosting us & Abigail Haddad for connecting us!

🌟 Exciting Event Alert! 🌟 We're hosting a not-to-be-missed talk with Chris Toth, Global Director of Sales Engineering at Hasura, who will be diving into the fascinating world of Supergraph technology. 🚀 What Will You Learn? The evolution of architectural trends, from monoliths to microservices. Insights into the burgeoning field of GraphQL-based data supergraphs. Practical knowledge that could influence your next project or career move. Event Details: 📆 Date: April 8th, 2024 🕒 Time: Noon EDT 🔗 Event Link Whether you're a developer, a data enthusiast, or just keen on the latest tech trends, this talk is for you. It's a fantastic opportunity to learn, network, and ask questions directly to an industry leader. Don't forget to mark your calendar! 📅 We're looking forward to seeing many of you there. #TechTalk #GraphQL #Supergraph #DataEngineersLunch #Microservices

Happy #OpenDataWeek! 🎉 #DidYouKnow DC's open data promotes transparency by offering over 1,400 curated datasets, maps, and apps? 📊🗺️ Check out the possibilities of #opendata today! Visit ➡️ opendata.dc.gov