— In the News —
This is awesome. This new "publication" offers an interactive ecosystem that lets readers engage directly with machine learning research. It's much more than a journal. This post in the Google Research Blog offers a good overview and definitely don't miss the journal itself.
Bringing machine learning products to market is increasingly a multi-disciplinary activity. If you're designing, managing, selling, or otherwise supporting machine learning products, this is a good starting point for what you need to know.
— Tools and Techniques —
You might not like where this goes but it sure is well done. This article by Trevor Martin uses a technique called latent semantic analysis to characterize activity on Reddit in a way that allows posts to be compared and quantified. This is a super interesting read.
Customer segmentation helps to identify different types of customers in your data. That can be useful to figure out which customers you might want more of and potentially, how to get them. This is a nice introduction to customer segmentation using K-Means clustering.
Graph analytics offers insights about the relationships between entities and interconnected objects. This step-by-step tutorial offers an overview of graph analytics, its uses, and how to get started with Python.
Transfer knowledge allows ML algorithms to transfer knowledge from one task to another. Andrew Ng has called it the next driver for machine learning commercial success. This article is a great overview of what transfer learning is, how it works, and applications.
— Resources —
The world's first linear algebra book with fully interactive figures.
— Data Viz —
Elijah Meeks is a Senior Data Visualization Engineer at Netflix, is the author of D3.js in Action and is one of the pillars of the data visualization community. Following a recent survey of over 1000 in the community, Elijah offers insights into the profession and where it's going.