Roundup 12/2/2016

freevector-futuristic-machinery

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

ZDNet has a special section on AI and machine learning. I’ve pulled some of the interesting pieces and linked them in the appropriate sections below.

Issues

— In Data Science Central, William Vorhies asks: has AI gone too far?  The context of his question is a paper that summarizes research into detecting criminality from facial images. In short, the researchers were able to successfully distinguish criminals from non-criminals in a sample of Chinese men aged 18 to 55 solely from facial measurements extracted from pictures. Vorhies notes that the research is rigorous, and, while the paper has evoked a chorus of criticism for its implications, critics have not yet identified a flaw in the methodology.

— In an article about fake news, Vincent Granville opines that it’s hard to detect fake news with machine learning because it’s hard to define fake news.

Fundamentals

— Alison DeNisco explains why AI and machine learning need to be part of your digital transformation strategy.

— Hope Reese lists five ways to get started implementing AI and ML.

Research

— MIT researchers develop a computational model of the human brain’s mechanism for face recognition.

— Cognitive scientist Joscha Bach ruminates on the elements of human intelligence we seem to be missing in AI.

Methods and Techniques

— In a podcast, Jon Bruner and Pete Skomoroch interview Richard Socher, chief scientist at Salesforce, and discuss how to make neural networks more accessible.

Software/Services

— Health tech startup Health Catalyst launches Healthcare.ai, a suite of open source packages for healthcare machine learning with R and Python APIs. The R package works with any R distribution and RStudio; the Python package works with Anaconda.

— Rescale’s Mark Whitney explains the ins and outs of running deep learning in the cloud. Rescale offers a managed service for deep learning in IBM Cloud.

— Reporting from AWS re:Invent, Doug Henschen argues that Amazon can make up for its late entry into machine learning services will be offset by its scale. File that under Things That Ain’t Necessarily So. AWS has a well-deserved reputation as the Stupid Cloud, and three services don’t begin to match what Microsoft, Google, and IBM offer.

— Nick Heath asks if Microsoft should be your AI and machine learning platform. He answers his own question by enumerating the many different services in the Cortana Intelligence Suite.

— Natalie Gagliordi asks the same question of Google.

— Hope Reese wonders if AWS should be your AI and machine learning platform. She quotes Gartner’s Alexander Linden, which is a bad sign.

— Conner Forrest asks if declining tech giant IBM should be your AI and machine learning platform. He doesn’t really answer the question.

Applications

— In Forbes, Suparna Goswami explains how an Indian startup uses machine learning for smarter hiring.

Companies

— Also in Forbes, Aaron Tilley chronicles NVIDIA’s transformation from a maker of gaming chips to a maker of AI chips.

— Jermy Hsu profiles Maluuba, a startup that uses deep learning to understand speech.

— Connor Forrest reports on five upstarts that are “leading the AI and machine learning revolution”: Uber, Tesla, Salesforce, NVIDIA, and Ayasdi. Wait, what? Ayasdi?  Also, for the record, Salesforce may be buying companies, but it’s not exactly leading the charge in machine learning.

Bottom Story of the Day

— GE CEO Jeff Immelt says he’s ready for Trump.

Roundup 12/1/2016

Simulationen – Anwendungen:
Was ist die optimale Schaufelform von Turbinen? Wie stark oszilliert der Druck in der Brennkammer? Computersimulationen liefern die Antwort.

Simulation - Applications:
What´s the optimum shape for a turbine blade? How much does the pressure in a combustion chamber oscillate? Computer simulations provide the answers.

PoF Frühjahr 2006, S. 81f			Siemens-Pressebild / Siemens press photo

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

Events

— Spark Summit East meets in Boston February 7-9 2017.

AWS Launches AI Services

Amazon Web Services launches three new services:

A massive medialanche ensues.

Serdar Yegualp applauds. In TechCrunch, Frederic Lardinois touts the announcement as Amazon sharing its “machine learning smarts”. Meh. It seems to me that AWS is well behind Google, Microsoft, and IBM in machine learning.

People

— In a podcast, Ben Lorica interviews Mike Franklin, co-director of Berkeley’s recently wrapped-up AMPLab project, who talks about AMPLab’s legacy. That legacy includes Spark, Alluxio, BlinkDB, KeystoneML, and Succinct, among other projects.

Fundamentals

— Aatash Shah explains the differences between machine learning and statistics.

Research

— In the second installment of a planned series on deep learning research, Adit Deshpande explains reinforcement learning. The first installment covered generative adversarial nets.

— Professor Neil Lawrence of the University of Sheffield suggests that you avoid panic, assures you that deep learning will be mostly harmless, and offers some thoughts on new directions in kernels and Gaussian processes. He adapted his blog post from a recent talk at a workshop aptly named New Directions for Learning with Kernels and Gaussian Processes.

Methods and Techniques

— In MIT Technology ReviewNicholas Diakopoulos and Sorelle Friedler propose a framework to ensure accountability for algorithms. They stress five principles: responsibility, explainability, accuracy, auditability, and fairness.

— On GitHub, Simon Brugman builds a collection of deep learning papers.

— Aaquib Saeed explains how to implement a convolutional neural network in TensorFlow for human activity recognition.

Software/Services

— The BigML team releases bigml 4.7.0, an open source Python binding to the public BigML API.

Applications

— Ram Shankar Siva Kumar describes an approach to testing security procedures by using machine learning to simulate attacks.

— Google uses machine learning to write the snippets that accompany search results.

— Charlie Osborne breathlessly explains how machine learning can stop terrorists from money laundering. She seems to think that machine learning in AML is a new thing.

— In The Huffington Post, Adi Gaskell describes how machine learning supports healthcare.

Companies

— In Fortune, Aaron Pressman speculates that big tech companies will compete to acquire machine learning companies in 2017. It’s a safe bet. After all, if Apple is willing to shell out $200 million for GraphLab Dato Turi, they’re willing to invest in anything that sounds remotely like machine learning.

Bottom Story of the Day

— The Internet Archive builds a replica database in Canada due to concerns about Donald Trump’s election. The irony there is that by locating the database in Canada, The Internet Archive will be subject to the Canadian PIPETA law, which includes EU-style restrictions on data collection and governance, including the “right to be forgotten.” So the Canadian version of the Wayback Machine may have to delete all those references to Max Mosley’s Nazi S&M sex scandal.

Roundup 11/30/2016

abstract-wallpaper-54

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

Events

— Spark Summit East meets in Boston February 7-9 2017.

Good Reads

— From Moor Insights & Strategy, an excellent paper on trends in predictive maintenance, a key application for machine learning. (Update: A reader notes that this is a sponsored paper, which is true, but I have no business relationships with the author or sponsor and thought it was a good read anyway.)

— In The Next Platform, Nicole Hemsoth explains the supercomputing vision of NVIDIA CEO Jen-Hsun Huang.

Scenes of the Future

An MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) deep learning system produces videos that predict what will happen next in a scene based on a single image from that scene. In ZDNet, Liam Tung reports. Linkapalooza here.

People

— Max Kuhn, author of Applied Predictive Modeling and progenitor of the caret package for machine learning moves to RStudio.

Methods and Techniques

— On Codementor.io, Matthew Corrigan introduces you to machine learning with Python and perceptrons.

Software/Services

— At AWS’ re:Invent Conference, Databricks announces HIPAA compliance for its Apache Spark managed service. Databricks has also achieved AWS Public Partner status.

— In ZDNet, George Anadiotis touts the latest release of Birst, a BI tool with some embedded ML capabilities. He mentions automated machine learning, but it’s clear that he has no idea what he’s talking about. Pro tip: if you think that BeyondCore is an example of automated machine learning, go to the back of the class.

— Splice Machine ships Release 2.5 of its eponymous product. The press release mentions machine learning but the product has no ML capability, so I guess it’s all about SEO.

— TensorFlow v0.12.0 RC0 is now available, and it runs on Microsoft Windows. Features available on Windows are a subset of the full feature set. For details, read the announcement.

— In Fortune, Barb Darrow explains why AWS has standardized on MXNet for deep learning.

— Fujitsu introduces consulting services to help customers accelerate their use of machine learning and AI.

Hardware

— In Next Big Future, Brian Wang swoons over the NVIDIA Xavier chip and its potential for deep learning.

Applications

— A startup named ebo uses a neural network to help you choose gifts for people. The service is still in preview. Michael Irving reports.

— GE Healthcare and Boston Children’s Hospital partner to develop deep learning tools for pediatric brain scans.

— In MIT Technology Review, Will Knight describes how a Google eye scanning algorithm can diagnose diabetic retinopathy better than human experts can. On the Google blog, Lily Peng explains. The JAMA paper is here.

— In Security Week, Kevin Townsend explains how machine learning helps criminals attack systems. Hey, machine learning doesn’t do bad things, people do bad things.

— On the Yelp Engineering blog, Alex M. describes how he used deep learning to find beautiful photos on Yelp.

Companies

— Sanghamitra Kar profiles Delhi-based Lybrate, a startup that uses ML for healthtech.

— Petuum Inc, a spinout from Carnegie Mellon University, lands $15 million in venture capital to “democratize” machine learning.

— The Cortana Intelligence and Machine Learning Blog explains how Microsoft shares machine learning and data science within the enterprise.

Bottom Story of the Day

— In The Eponymous Pickle, Franz Dill reports on sex as an algorithm.

Roundup 11/29/2016

big_thumb_a5b37d326cd1aabe1dc4ee5f7988d829

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

Events

— Spark Summit East will meet in Boston February 7-9 2017.

Good Reads

— In The Next Platform, Nicole Hemsoth explains why Microsoft invests in FPGAs for compute-intensive applications like machine learning. Separately, Nicole investigates Intel’s strategy to integrate the deep learning assets it acquired when it bought Nervana earlier this year.

Issues

— Benedict Evans speculates about the impact of cheap and pervasive cameras combined with image recognition technology.

— The Economist claims that economists love fads and machine learning is the latest fashion. The article is self-debunking. The chart below shows gradual assimilation over thirty years, not Pavlovian fashion-following.

20161126_fnc301

Fundamentals

— In Harvard Business Review, Anastassia Fedyk explains how to tell if machine learning can solve your business problem.

— Tyler Lacoma argues, cryptically, that machine learning is more advanced than ever before, but it’s not Judgement Day yet. That’s a relief.

— A Quora reader requests advice for novice machine learning users who feel overwhelmed. Sebastian Raschka responds. My advice: change careers. Successful data scientists tend to feel energized by all of the resources available today.

— Adrian Sampson describes three common statistical mistakes and how to avoid them.

Methods and Techniques

— Here is the complete series of posts on Topic Modeling from the BigML blog. If you don’t know what Topic Modeling is, read the series.

— Bioinformatics maven Shirin Glander asks: can we predict flu deaths with machine learning and R? She proceeds to answer the question by demonstrating multiple ways to do so in a tour de force post, with graphics and code snippets.

unnamed-chunk-4-1

Software/Services

— Serdar Yegualp explains why AWS standardized on MXNet for DL.

— Baidu releases “Long Utterance,” a set of Chinese language APIs for its speech recognition technologies.

— The Editorial Team at Inside Big Data gushes over IBM Watson Machine Learning because it is “built on Apache Spark.”  In fact, IBM Watson Machine Learning is a rebranded managed service for SPSS Modeler.

Hardware

— TechCrunch reports that Japan’s Ministry of Economy, Trade, and Industry plans to spend 20 billion yen to build a supercomputer capable of 130 petaflops. That’s a lot of yen and a lot of flops.

Applications

— NVIDIA Foundation awards $200K to the Translational Genomics Research Institute to perfect its software for the analysis of cancer cell genes.

— Nokia offers embedded machine learning to telcos for mobile customer experience analytics and customer care.

— In Business Insider, Lydia Ramsey describes five impactful innovations in radiology; machine learning drives two of the five.

Companies

— The MIT engineers and Wall Street analysts at Trefis see autonomous vehicles driving demand for NVIDIA’s DRIVE PX-2 deep learning platform. The same team explains NVIDIA’s steady margin growth.

Bottom Story of the Day

— In TechCrunch, Chris Nicholson argues that machine learning can fix America.

Roundup 11/28/2016

gears_machinery_desktop_1582x1214_hd-wallpaper-534925

ICYMI: Top machine learning (ML) and deep learning (DL) stories from last week.

News

— Google announces a $4.5 million investment in the Montreal Institute for Learning Algorithms. PR avalanche ensues.

— In Investopedia, Richard Saintvilus chronicles the cloud machine learning wars.

— Six months after releasing DSSTNE, an internally developed deep learning framework, AWS announces plans to standardize on MXNet. In Fortune, Barb Darrow reports. It seems like a smart move. MXNet is much more mature than DSSTNE, with a 10X larger contributor base. So I guess this move consigns DSSTNE to the dustbin of history.

Good Reads

— Adrian Colyer summarizes four papers:

  • A report from Stanford University covering key topics in artificial intelligence, including large scale machine learning, deep learning, reinforcement learning, robotics, computer vision, natural language processing, algorithmic game theory, IoT, and neuromorphic computing.
  • Training AI to play a first-person-shooter (FPS) game with deep reinforcement learning.
  • Sequence learning in Google’s Smart Reply feature.
  • Building machines that think and learn like people.

— On the Moor Insights & Strategy blog, highlights from SC16.

Top 20 Python ML Projects

In KDnuggets, Prasad Pore profiles the top twenty open source projects for machine learning with Python. His selection criteria are unclear; he includes TensorFlow, for example, because it has a Python API, but that means Spark ML, H2O and Microsoft Cognitive (CNTK) should make the list as well.

Also, Pore ranks projects by cumulative lifetime commits, a measure that correlates with the age of the project; commits in the past year is a better measure of current activity. Had he used that measure, the top projects would be Spark, CNTK, H2O, TensorFlow, and Theano, in that order, with scikit-learn a distant sixth and Caffe barely registering.

Aside from using an inappropriate measure and excluding important projects, it’s a fine piece of analysis.

python-top20-2016-bubble-chart

Explainers

— In Fortune, Aaron Pressman argues that Intel’s strategy for machine learning and AI is smart, but lags NVIDIA. Nicole Hemsoth describes Intel’s approach as “war on GPUs.”

— Timothy Prickett Morgan dissects the “Summit” supercomputer on order from the U.S. Department of Energy for its Oak Ridge National Laboratory.

— In Wired, Cade Metz profiles Intel’s chip strategy for machine learning and deep learning. In EETimes, Rick Merritt explains how Intel’s Nervana attacks GPUs.

— Adrian Colyer summarizes Microsoft’s recent paper documenting how they built an application that recognizes conversational speech as well as humans do.

— Google Brain’s Mike Shuster (and two other Googlers) explain “zero-shot translation,” or the ability to translate between language pairs that the system has not seen before.

— Remember the glowing reports about the Clinton campaign’s competitive advantage in advanced analytics and machine learning? Like this and this and this and this and this? As it turns out, the Trump campaign had an advanced analytics operation, too. In the future, data science won’t bestow a competitive advantage to one side or the other in political campaigns. It will be table stakes.

— Sebastian Raschka, the author of Python Machine Learning, opines on programming languages for machine learning. Guess which language he favors.

Roundup 11/23/2016

invented-first-supercomputer_ae3185a5bd50219

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

Note to readers: owing to the U.S. Thanksgiving holiday, I do not plan to publish a Roundup on Thursday, November 24 or Friday, November 25. 

AWS Chooses MXNet

Six months after releasing DSSTNE, an internally developed deep learning framework, AWS announces plans to standardize on MXNet. In Fortune, Barb Darrow reports. It seems like a smart move. MXNet is much more mature than DSSTNE, with a 10X larger contributor base. So I guess this move consigns DSSTNE to the dustbin of history.

Good Reads

— Still cleaning up last week’s hot mess of announcements. In Wired, Cade Metz profiles Intel’s chip strategy for machine learning and deep learning. In EETimes, Rick Merritt explains how Intel’s Nervana attacks GPUs.

— On his eponymous blog Elad Gil speculates that Facebook sucks at machine learning.

— Adrian Colyer summarizes Microsoft’s recent paper documenting how they built an application that recognizes conversational speech as well as humans do.

Fundamentals

— Sandeep Raut asks and answers: what is deep learning?

— In Lifehacker, of all places, Eric Ravenscraft explains what neural networks, artificial intelligence and machine learning do in your mobile apps.

Research

— Google Brain’s Mike Shuster (and two other Googlers) explain “zero-shot translation,” or the ability to translate between language pairs that the system has not seen before.

Methods and Techniques

— In DZone, Sibanjan Das explains how to use Spark ML pipelines.

— On the Google Cloud Big Data and Machine Learning Blog, Justyn Kestelyn explains how to use real-time visualization and machine learning to understand urban traffic flows. He uses London as an example.

— Jason Brownlee offers a cheat sheet for improving the performance of machine learning techniques.

Software and Services

— On the main Google Blog, Cindy Teruya touts nine ways the Google Cloud Machine Learning can help businesses.

— Tom Radcliffe compares Python and R for machine learning. He leans toward Python.

— For some reason, Susan May revisits the tiresome old Spark versus MapReduce question. She provides a useful summary of why it’s time to stick a fork in MapReduce — it’s done.

Hardware

— A blogger on Seeking Alpha argues that Intel shouldn’t go to war against NVIDIA, and should offer its own GPUs instead.

Applications

— Remember the glowing reports about the Clinton campaign’s competitive advantage in advanced analytics and machine learning? Like this and this and this and this and this? As it turns out, the Trump campaign had an advanced analytics operation, too. In the future, data science won’t bestow a competitive advantage to one side or the other in political campaigns. It will be table stakes.

Companies

— In Forbes, Jon Markman gushes over NVIDIA’s third quarter earnings report. So does everyone else on Wall Street, it seems, judging by this linkapalooza.

Bottom Story of the Day

— IBM Watson expands consulting support to the Internet of Things.

Roundup 11/22/2016

maxresdefault

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

Events

— Spark Summit East will meet in Boston February 7-9 2017.

Good Reads

— Adrian Colyer summarizes a report from Stanford University covering key topics in artificial intelligence, including large scale machine learning, deep learning, reinforcement learning, robotics, computer vision, natural language processing, algorithmic game theory, IoT, and neuromorphic computing.

— On the Moor Insights & Strategy blog, highlights from last week’s Supercomputing conference.

Fundamentals

— In MIT Technology Review, Will Knight describes some experiments recently released by Google that demonstrate how neural networks work.

— Zachary Chase Lipton writes a long thumb-sucker about algorithmic bias.

Methods and Techniques

— On the BigML blog, “talvarez” explains how to predict movie review sentiment with topic models. Part four of BigML’s Topic Modeling series is here.

— Don Hillborn explains oil and gas asset optimization with Amazon Kinesis, Amazon RDS, and Databricks.

— Engineers develop a new machine learning algorithm that learns from human instruction and not from data. That strikes me as an oxymoron.

Hardware

— In Fortune, Aaron Pressman argues that Intel’s strategy for machine learning and AI is smart, but lags NVIDIA. Nicole Hemsoth describes Intel’s approach as “war on GPUs.”

— Jelor Gallego reveals that NVIDIA just built the most energy-efficient supercomputer ever.

— Timothy Prickett Morgan dissects the “Summit” supercomputer on order from the U.S. Department of Energy for its Oak Ridge National Laboratory.

Applications

— Sophie Curtis asks: how can machine learning create a smarter energy grid?

— On the WeWork blog, Nicole Phelan describes how WeWork uses machine learning to design offices. A neural network significantly outperforms human designers at predicting utilization.

Companies

— A blogger on Seeking Alpha argues for a lot of upside in NVIDIA shares.

— Serkan Piantino, co-founder of Facebook’s AI research lab, quits to start a new venture, Top 1 Networks, which will provide GPU-accelerated computing as a service.

— Google announces a $4.5 million investment in the Montreal Institute for Learning Algorithms. PR avalanche ensues.

— In Investopedia, Richard Saintvilus chronicles the cloud machine learning wars.

Bottom Story of the Day

— Pinterest uses machine learning to determine what’s trending.

« Older Entries