The Harvard Business Review has been taking a hard look at the practical side of AI for businesses recently. This latest article explores the capabilities that should already be in place before thinking about adding AI. Like it or not, you shouldn't ignore the basics.
Great idea. In this crowdsourced analytics project, many teams will analyze the same data and then compare the code and results. In this phase, they're trying to pinpoint exactly why analytic choices have such a profound effect on research results. This Google Doc offers details and instructions for getting involved.
Python is the ideal language for data science, but getting set up with all the libraries you need can be time-consuming. ActivePython is pre-bundled with over 300 packages including NumPy, SciPy, scikit-learn, TensorFlow, Theano and Keras, and is integrated with the Intel Math Kernel Library (MKL) for optimized NumPy and SciPy computations. It’s free to use in development, so you can get started in minutes.
Google just released a new feature to their free spreadsheet product that enables users to simply ask questions of their data using natural language. For instance, you can ask “what is the distribution of products sold?” or “what are average sales on Sundays?” and Explore will help you find the answers.
Ryan Dahl is a very accomplished software engineer who got to spend a year in the Google Brain Residency Program. In this post, Ryan takes a look at some of the projects he worked on and describes what it's like to dive into machine learning from an engineering perspective. His Thoughts and Conclusions section, especially, is gold.
Loops aren't typical in Deep Learning systems but researchers are discovering that loops are creating mind-boggling capabilities for automation. This overview explores how loops work in a variety of network architectures and why "strange loops" are the fundamental reason for what Yann LeCun describes as the coolest idea in machine learning in the last twenty years.