ML/AI

Machine learning (ML) and deep learning (DL) content from the past 24 hours.

Google Enhances Machine Learning Services

Google launches a dedicated team to drive Google Cloud Machine Learning, and announces a slew of enhancements to its services:

— Google Cloud Jobs API provides businesses with capabilities to find, match and recommend jobs to candidates. Currently available in a limited alpha.

— In 2017, GPU-accelerated instances will be available for the Google Compute Engine and Google Cloud Machine Learning. Details here.

— Cloud Vision API now runs on Google’s custom Tensor Processing Units; prices reduced by 80%.

— Cloud Translation API will be available in two editions, Standard and Premium.

— Cloud Natural Language API graduates to general availability.

Infoworld’s Serdar Yegualp is impressed. Linkapalooza here.

Microsoft Partners for AI

— Not-for-profit AI research organization OpenAI announces a partnership with Microsoft to advance AI research. Microsoft Azure will serve as OpenAI’s primary cloud platform. Linkapalooza here.

— Separately, Microsoft and NVIDIA announce a partnership to accelerate AI. Once again, Microsoft Azure will serve as the hub. Linkapalooza here.

Fundamentals

— In Forbes, investor Jason Black offers a hype-free explanation of machine learning.

— Sam DeBrule provides a non-technical guide to machine learning and artificial intelligence.

— Camron Godbout delivers a simple description of deep learning.

Methods and Techniques

— In O’Reilly Radar, Reza Zadeh explains why deep learning is hard. In a word: optimization.

— Separately, S. Zayd Enam explains why machine learning is hard.

— Meanwhile, Matthew Honnibal proposes a four-step approach that makes natural language processing with deep learning easy.

— On the BigML blog, someone who blogs under the username “mariajesusbigml” introduces you to Topic Models.

Software/Services

— Martin Heller reviews Spark ML 2.01. He likes it.

— Separately, Martin reviews Microsoft Cognitive Toolkit (CNTK). He likes that, too. Looking back at Martin’s reviews of TensorFlow, Google Cloud Machine Learning, IBM Watson, and Amazon Machine Learning, I gather that Martin likes everything, except for HPE Haven, of course, but that’s a given. Oddly, Martin thinks Microsoft Azure Machine Learning is harder to use than the rest of this lot, which is nuts.

— Cray unveils deep learning toolkits for the Cray XC and Cray CS-Storm systems. Kits include Caffe, Microsoft Cognitive Toolkit (CNTK), MXNet, NVIDIA DIGITS (Deep Learning GPU Traning System), TensorFlow, and Torch.

Applications

— In Fast Company, Christina Farr describes a joint project of General Electric and the University of California – San Francisco to develop a library of machine learning algorithms for diagnostics in primary care, pathology, and radiology.

— Nicole Hemsoth describes the Cancer Distributed Learning Environment (CANDLE) framework, powered by NVIDIA. It’s part of the “Cancer Moonshot” initiative. More stories on this here.

— Scientists at UCLA and the University of Illinois use machine learning to discover and design alpha-helical membrane-active peptides, which may lead to improved cancer therapies.

Companies

— In The Motley Fool, Timothy Green argues that NVIDIA is crushing AMD.

— DataRobot, a machine learning company, partners with data blending leader Alteryx.