— Insight —
Over-hyped claims about AI have contributed to past AI winters and, as Gary Marcus explores in this article, we may be headed down that same path again. Here's what we can do to stop it.
Privacy or data? It's a complicated relationship but it doesn't have to be one or the other if the right compromises are made.
— Tools and Techniques —
Nbdev brings the key benefits of IDE development into the notebook system so you can continue to work in notebooks when you're ready to move on from exploratory development. This new open-source tool by Jeremy Howard and Sylvain Gugger offers features like documentation lookup, good syntax highlighting, integration with unit tests, and the ability to produce distributable source code files - all within Jupyter.
There aren't implementation details here but it's a good think-through for building an automated trading system. Your results may vary but the author claims one "4-month period without a single losing day." The Hacker News discussion is also worthwhile.
Nice introduction to Kubernetes, starting with the problems it solves, what it is, and how to use it.
In this visual introduction to BERT, Jay Alammar shows how to use a variant of BERT to classify sentences. This is an example that is basic enough for an initial introduction, yet advanced enough to showcase some of the key concepts involved.
Dolt combines the convenience and ease of use of a relational database, with the elegance of the Git version control model, all delivered as an open source tool.
— Resources —
Looking for data? This collection of publicly available datasets covers a wide range of domains and potential uses.
— Data Viz —
When two people look at the same visualization, they don't necessarily see the same thing. People tend to see what they expect to see and that can be especially problematic when creating visualizations for experts. This post by Cindy Xiong shows the problem with ideas for handling it.
— Conferences & Events —
- Metis Webinar | AI ROI: The Questions You Need to Be Asking - As business leaders increase investment in advanced analytics, data science, and AI, many struggle to recognize a return on those efforts. During this free Metis Corporate Training webinar Kerstin Frailey, Senior Data Scientist and Head of Executive Corporate Training at Metis, will walk through what you need to ask before, during, and after the lifetime of a data science project. Thursday, December 5, 12pm ET
- Experiment Management: Rethink your Machine Learning Workflow - Join Comet.ml data scientist, Niko Laskaris, for a free webinar to learn how experiment management platforms help data scientists and teams track, compare, explain and reproduce their ML experiments leading to improved team collaboration, productivity and visibility. Tuesday, December 10, 2pm ET Register here >>