— In the News —
When Matt Chapman sent FOIA requests for metadata to agencies throughout the U.S., he expected agencies to push back. And they mostly did. But Seattle? Now there's a story...
— Tools and Techniques —
In the final article of this series, Sebastian Raschka compares the performance of machine learning models and algorithms using statistical tests and nested cross-validation.
This post from the engineering blog at Uber shows why you can't judge feature changes by the average customer experience. Analyzing distributional changes are far more effective. Here's why.
Great interactive tutorial that introduces tidyeval principles and tools. This is well organized and includes inline, runnable code examples.
It seems like everybody wants to build their own custom machine learning stack these days. Many organizations have good use-cases but where they get stuck is actually making it work, hiring people, and making them successful. In this short post, Lukas Biewald, CEO of Weights & Biases, suggests 3 approaches, based on the size of your team.
Being able to give a meaningful and engaging talk can be crucial for getting support for your work but most people don't get a lot of practice. Here are key tips for designing visuals, deciding what to include, and ways to deliver your message. Includes linked references and examples of noteworthy data presentations.
Increase your chances of success with extra practice. Pramp is the world’s largest peer-to-peer mock interviewing platform. Now you can practice not only DS&A, system design, and behavioral interviews but also data science interviews. Pramp - a peer-to-peer mock interviews platform, where you can practice your interviewing skills on demand for FREE. Schedule a free practice →
— Resources —
This curated collection of machine learning resources includes sections for open courses, free books, tutorials, topic-specific lecture notes, and cheat-sheets. This is well organized and got a lot of attention around the web this week.
If you're interested in reinforcement learning, you'll definitely want to check out this new educational resource from OpenAI. Spinning Up in Deep RL consists of crystal-clear examples of RL code, educational exercises, documentation, and tutorials.
— Career —
Don't try to be a jack-of-all-trades. What kind of data scientist do you want to be?