Data Science Roundup #67: Focus on Deep Learning! Plus: The 10 Best Data Viz Projects of 2016

 Over the past few days, Google’s Deepmind machine-learning team secretively put its AlphaGo artificial intelligence system onto two Chinese online board-game platforms to test its skill in fast-paced games against several of the world’s best Go players. It won every game it played. Go has become the province of AI, and DeepMind further proves that GANs are an extremely promising approach.

This is the best summary of deep learning activity over the past year that I’ve read. It focuses on new techniques (reinforcement learning and adversarial networks), the trend towards openness in the commercial research space, the white-hot acquisition market, and the advances in chips. 

Very readable; highly recommended if you don’t follow deep learning closely.

While the title of this post says “non-comprehensive”, that’s only really true in the most technical sense. This is a massive list of (38!) awesome developments in the world of stats in 2016. 

Assuming you like lists of awesome things, this is an absolute must read.

Running a regression is a commonplace task—you can do it via a function call it any statistical programming language (even Excel!). But do you really understand what that function is doing? I always feel far more confident using functions when I understand the details of their implementation, and I love this post for diving into the internals of lm(). By the end, you could implement linear regression from scratch.

Fishtown Analytics works with venture-funded startups to implement Redshift, BigQuery, Mode Analytics, and Looker. Want advanced analytics without needing to hire an entire data team? Let’s chat.

Developers shouldn’t have to write ETL scripts. Consolidate your data in minutes. No API maintenance, scripting, cron jobs, or JSON wrangling required.