ISSUE 442 · June 27, 2023
Note that Data Elixir is taking next week off and will be back in your Inbox in two weeks. If you're in the U.S., have a great Fourth!
Insight
The economic potential of generative AI: The next productivity frontier
New tools, articles and posts about generative AI seem to be everywhere these days but ultimately, the era of generative AI is just beginning. This report from McKinsey explores this burgeoning new space, including business opportunities, impacts on industry, implications for workers, and big-picture societal considerations.
McKinsey & Company
Sponsored Link
Deepchecks Revolutionizes ML Monitoring with Open-Source Release!
Deepchecks, known for their popular open-source test suites (650,000+ downloads), now introduces their open-source ML monitoring solution. Experience the future of collaborative AI/ML validation and empower your models!
Posts & Tutorials
LLM Powered Autonomous Agents
Great introduction to building autonomous agents using LLMs. In an LLM-powered autonomous agent system, the LLM functions as the agent’s brain and is complemented by components for planning, memory, and tool use. The post explores strategies, methodologies, and algorithms for each component with lots of references along the way.
Lilian Weng
Order Constraints in Bayes Models
In Bayesian modeling, order constraints refer to restrictions imposed on the relationships between a model’s parameters. These constraints help to incorporate prior knowledge and assumptions into the model. This tutorial walks through a practical example and shows how order constraints are useful and how to use them using the {brms} package.
Mattan S. Ben-Shachar
Get better at Data Science every week
Join 25,000+ aspiring data scientists and receive actionable Python & Data Science tips every Tuesday. Sign up for free!
// sponsored
Tools & Code
Mosaic
Mosaic is an open-source framework for linking data visualizations, tables, input widgets, and other data-driven components, while leveraging a database for scalable processing. Use it to interactively explore massive datasets, build data-driven web apps, or interact with data directly in Jupyter notebooks.
GitHub | UW Interactive Data Lab
WizMap - Interactive Viz for ML Embeddings
WizMap is an interactive visualization tool for exploring embeddings. It makes it easy to explore millions of points and get insights from multi-resolution summaries. Follow the links for a live demo and paper.
GitHub | Polo Club of Data Science at Georgia Tech
aeon
aeon is an open source toolkit for learning from time series compatible with scikit-learn. It provides access to the very latest algorithms for time series machine learning, in addition to a range of classical techniques for learning tasks such as forecasting and classification.
GitHub | aeon-toolkit
Resources
polars cookbook
This cookbook is a fork of the popular pandas-cookbook and has been modified to use the polars library instead of pandas. It uses real-world examples with "all the bugs and weirdness that entails."
GitHub | Escobar West
ML system design
How do companies like Netflix, Airbnb, and Doordash apply machine learning to improve their products and processes? In this collection of 200 case studies, 64 companies share practical use-cases and learnings from their machine learning systems.
Evidently AI
 |