Machine learning (ML) and deep learning (DL) content from the past 24 hours. Plus, some AI stuff.
— In Forbes, Adrian Bridgewater asks whether Microsoft can decode the future of artificial intelligence. I’m guessing that it can, but will first make you upgrade to Windows 10.
— Ian Barker explains deep learning.
— Jason Brownlee contemplates randomness as if he discovered it, or something.
Methods and Techniques
— Jean-Francois Puget argues that ML is actually mathematical optimization, which seems accurate. What’s interesting is that there are so few optimization algorithms in use: stochastic gradient descent, conjugate gradient, L-BFGS and a few others. Master optimization and you master ML.
— Joseph Chen asks how Lean Six Sigma can help machine learning. It certainly can’t hurt.
— Joe Rickert explains the Naive Bayes classifier.
Software and Services
— Researchers at Princeton and the University of California use ML to identify suspicious domain names.
— John Pitcher discusses the potential for ML to control building systems, such as HVAC.
— Symantec introduces a new release of Endpoint Protection, which uses ML to defeat cyber threats.