Roundup 10/28/2016


Machine learning (ML) and deep learning (DL) content from the past 24 hours. Plus, some AI stuff.

I’m publishing a three-part series on the state of enterprise machine learning in The Next Platform. Part one is here.

Good Reads

— Helen Beers explains AI in the second part of a series. Part one is here.

— Market Realist publishes a twelve-part series on the outlook for NVIDIA. Don’t worry, they’re short parts.


— In Wired, Clive Thompson whines that AI isn’t accountable, citing an example where a consumer’s application for insurance is declined without explanation. It’s pretty clear that Clive Thompson has no idea what he’s writing about:

  • AI-driven applications are accountable to the people who build them. Banks and insurance companies don’t want to decline your business; they want to approve your application unless you’re a bad risk.
  • A biased model is bad for business, and the data scientists who build them know it. That’s why they constantly monitor models for systematic errors.
  • Any predictive model, including “black box” learners, can be designed to deliver explanations for rejects. In regulated industries like banking and insurance, this is legally required.

Other than that, it’s a fine article.

Methods and Techniques

— Researchers at Google Brain publish a paper on how neural networks can learn to use secret keys to protect information when they communicate with one another.

— Daniel Tunkelang, formerly of LinkedIn and Google, is skeptical that we can automate ML.

— If you’re inclined to feel that convolutional neural networks get way too much buzz and recurrent neural networks are underappreciated, William Vorhies is on your side.



— Kaggle launches a product recommendation challenge for Santander Bank. Submissions due 11:59pm Wednesday, 21 December UTC. First place gets $30K; second place gets $20K; third place gets $10K; fourth place gets a set of steak knives.


— Twitter says it will use ML to make your timeline more relevant. No word on whether they plan to use ML to decide who to fire.

— Netradyne announces three new DL-driven applications for its Drivieri fleet safety platform: traffic light detection, relative speed determination, and pedestrian identification.

— To defend against all of those lawsuits we can expect from the introduction of autonomous vehicles, Intraspexion introduces a DL-driven application to identify potential litigation risks.

— IBM adds “cognitive behavioral biometrics” to its digital banking fraud prevention technology. That’s where they measure how you interact with the keyboard and mouse to create a biometric profile.

— Datadog introduces ML-based anomaly detection to help engineering teams detect abnormal behavior in a cloud environment. When the software detects unusual behavior, it barks. Just kidding.

— Parker Beauchamp surveys AI applications in the insurance industry.

— Purdue researchers develop a DL-driven application to assess disaster damage from images.

— Yahoo Tech reports on DL-driven robots that cut meat.

ML/DL Software and Services

— Databricks adds support for DL with TensorFrames, a Spark package that enables TensorFlow to run on Spark. The managed service includes support for TensorFrames with GPU-accelerated compute instances. Alex Woodie reports, noting that Spark ML runs 10X faster on GPUs.

Hardware and Provisioning

— Scientific Computing World reports on the challenges of managing an HPC cluster for applications like deep learning.

— Rich Brueckner asks if deep learning will scale to supercomputers. The answer seems to be “yes”.

Roundup 10/27/2016


Machine learning (ML) and deep learning (DL) content from the past 24 hours. Plus, some AI stuff.

I’m publishing a three-part series on the state of enterprise machine learning in The Next Platform. Part one is here.

The latest Kaggle newsletter is here, with lots of interesting content.

Good Reads

— Andrew Colyer summarizes a paper on the use of graph analytics to detect system intrusions at an early stage (before the hackers can do any harm.)


— Ben Dickson fears the “dark side” of machine learning.

— Stephen Gardner summarizes data privacy challenges posed by AI.

Marketing and Ad Tech

— Adgorithms, an ML-driven ad tech startup, uses AI to drive sales for Harley-Davidson.

— Qubit launches Qubit ML, an ML-driven digital experience management platform.


— Forbes asks: what product breakthroughs will recent advances in deep learning enable? Google Brain’s Eric Jang answers:

  • Customized data compression
  • Compressive sensing
  • Data-driven sensor calibration
  • Offline AI
  • Human-computer interaction
  • Gaming
  • Artistic assistants
  • Unstructured data mining
  • Voice synthesis

— Circadence, a startup that provides cyber security training, announces plans to use the NVIDIA DGX-1 system in Project Ares, a cyber security training platform.

— Reuters reports on the role of ML and AI in financial market surveillance.

— Adi Gaskell describes the work of two startups, Cogito and Canary Speech, who use speech recognition to assist medical professionals working with dementia patients.

— In CFO, Keith Button summarizes six examples of firms using ML for prediction. Also in CFO, Viral Chawda describes five ways to implement advanced analytics.

— GM plans to use IBM Watson to profile you and sell you stuff in your car.


— Dave Ramel describes Microsoft’s Cognitive Toolkit and IBM Watson Data Platform.


— Penguin announces the availability of two new Open Compute servers with NVIDIA GPU accelerators for deep learning.


— CaliberMind raises $1.1 million of seed capital. The company uses machine learning and language analysis to build psychographic profiles of buyers for B2B Sales and Marketing.

Roundup 10/26/2016


Machine learning (ML) and deep learning (DL) content from the past 24 hours. Plus, some AI stuff.

In Scientific American, Larry Greenemeier reports that AI is not out to get us. Well, that’s a relief.

Microsoft Launches Cognitive Toolkit 2.0

— MSFT rebrands CNTK as Cognitive Toolkit, releases version 2.0 to beta. Enhancements include a Python API, Visual Studio support, reinforcement learning and performance improvements. Linkapalooza here.



— At a Wall Street Journal Tech Conference, NVIDIA CEO Jen-Hsun Huang argues that technology for autonomous cars is evolving “faster than Moore’s Law.”

— I don’t usually link sponsored content, but this piece from NVIDIA is a good read.

— Steve Hanley describes the NVIDIA PX-2, which will provide Tesla with the computing power necessary for Level 5 autonomous driving. Matt Pressman provides additional detail.


— Alex Woodie reports on a Kaggle competition sponsored by Melbourne University, Mathworks, the American Epilepsy Society and the National Institutes of Health. The goal of the competition is to predict seizures in long-term intracranial EEG recordings. First place prize is $10K; the entry deadline is November 14.


— Ryan Joe explains AI to Marketers without using math.

— Helen Beers explains ML and AI.

Methods and Techniques

— Two how-to-do-its from Data Science Central:

— Disney Research reports a new method that reduces the amount of training data needed for facial performance capture in film and video game production.

— Stephen Max Patterson describes how robots can teach each other by exchanging data.


— Aussies deploy bots to control hunt down and kill crown-of-thorns starfish before they eat the Great Barrier Reef.

— In HBR, Ashby Rowe et. al. describe how ML-driven startups may transform the investment industry.

— John McCrank surveys the role of AI in regulating financial markets.

— In Slator, Florian Faes profiles some data scientists working on machine translation.

— Interesting list of ten deep learning apps investors should watch.

— Karl Zimmermann argues that machine learning may be the solution to enterprise cyber security problems.


— IBM announces YAWBSP (Yet Another Watson Branded Software Product), launches media blitz.


— Computer vision startup Movidius, which is about to be acquired by Intel, partners with China’s Hikvision to deliver intelligent surveillance cameras. Linkapalooza here.

— To encourage AI research, Facebook donates 22 GPU-accelerated servers to universities across Europe.

— Philips Lighting and ABB invest $7 million in PointGrab, an ML-driven “smart buildings” startup.

— Israeli startup Anodot launches a real-time incident detection service for fintech companies.

— Dataiku snags a $14 million “A” round.

— Image recognition startup Clarifai lands $30 million in funding.

Roundup 10/25/2016


Machine learning (ML) and deep learning (DL) content from the past 24 hours. Plus, some AI stuff.

In TechCrunch, John Mannes asks: WTF is machine learning? 1,342 words later, he still doesn’t know.

Methods and Techniques

— Researchers from Google Brain, Penn State, and OpenAI describe a method for semisupervised knowledge transfer from a “teaching” model to a “student” model. The “student” model never sees the original data, and cannot leak it accidentally. Dave Gershgorn reports.

— On the MapR blog, Carol McDonald explains how to use Spark’s logistic regression learner to predict cancer malignancy. She uses the Wisconsin Diagnostic Breast Cancer data set to build her tutorial.


— In the New York Times, John Markoff describes the criminal potential of AI. So, if you’re a crook, read this article.

Health and Medicine

— Lisa Cornish reports on six technologies that will disrupt health care in developing markets. Guess what’s on top of the list.

— AliveCor and Mayo Clinic announce that they will collaborate on a mobile electrocardiogram program designed to detect hidden health signals. In Recode, Kara Swisher reports.


— In Forbes, Peter High explains Gartner’s picks for the top ten technology trends for 2017, of which AI and machine learning rank #1.

— Knowlton Thomas summarizes how Google, Apple, Microsoft, Amazon, Salesforce, IBM and Intel use AI. He misses a lot.

— Tod Newcombe reports on four developing technologies, including AI, that could help solve modern problems. That seems a wee bit exaggerated.


— Folks at the MIT Media Lab develop a Nightmare Machine that uses AI to scare us. Gizmodo, BoingBoing, and The Atlantic all report.

— In Digiday, Ross Benes explains why newsrooms are expanding their data teams. For starters, they need people to report on stories like MIT’s Nightmare Machine.

— Ian Lopez describes four technologies (including ML) that are disrupting legal tech.

— Matthew Finnegan explains why machine learning will play a key role in data center operations. For example, Google uses DeepMind to manage power consumption at its server farms.

— MIT Technology Review reports on Microsoft’s accomplishment matching humans in conversational speech recognition, and why that matters.

— Jack Vaughan describes how Reltio Cloud uses graph analytics for master data management.

— Researchers at University College London develop an algorithm that predicts the outcome of human rights trials with 79% accuracy.  While that may sound impressive, bear in mind that if 79% of all trials result in a guilty verdict, an algorithm that declares all defendants to be guilty will be correct 79% of the time. We judge predictive models by how much they improve accuracy over alternative methods and not their absolute accuracy.


— Nozomi Networks announces the latest release of SCADAguardian, it’s ML-driven cyber security software.


— Paxata raises $33.5 million, plans to enhance the machine learning and semantic analysis capabilities of its data integration platform.

— NetSpeed Systems lands a $10 million “C” round to fund enhancements to its ML-driven System-on-a-Chip (SoC) design technology.

— In VentureBeat, Jordan Novet reports that NVIDIA sees government as its next goldmine. I don’t see why not; everyone else does.

Roundup 10/24/2016


Top machine learning (ML) and deep learning (DL) stories from last week, plus new content from Friday and the weekend.

The theme for featured images this week is art produced by deep learning.

ICYMI: Top Stories of Last Week

— AMD, Dell EMC, Google, HPE, IBM, Mellanox, Micron, NVIDIA, and Xilinx launch the OpenCAPI Consortium and industry group to promote specs for the next generation of data center hardware.


— Apple hires Carnegie Mellon University professor Ruslan Salakhutdinov as Director of AI research. Linkapalooza here.

— Andrew Oliver proposes dropping seven technologies from the Big Data ecosystem: MapReduce, Storm, Pig, Java, Tez, Oozie, and Flume. He forgets to mention Mahout, which is forgivable since nobody uses it.

— Daniel Gutierrez interviews Jim McHgh of NVIDIA’s Deep Learning Group, who says he wants to collaborate with Databricks to integrate the BIDMach machine learning library with Spark.

— Meanwhile, Gartner announces the top ten strategic technology trends for 2017, and machine learning is right up there at #1 on the list.

— Serdar Yegualp describes Microsoft’s big bet on FPGAs, explains the potential of FPGAs for machine learning, notes that existing machine learning software generally does not support FPGA acceleration.

— Meanwhile, however, Baidu announces that it will accelerate its machine learning applications with Xilinx FPGAs.

— Xilinx is on a roll. TeraDeep announces a fast deep learning solution that leverages Xilinx FPGAs.

— Using CNTK, MSFT researchers achieve parity with humans in speech recognition; medialanche ensues.

— In HBR, Tom Davenport explains how to introduce AI into your organization. The next generation of AI will introduce itself.

— Tesla announces that its new cars will include all of the hardware needed for level 5 autonomy. The software isn’t available yet but will be added through over-the-air updates.

Good Reads from Last Week

— Christine Barton et. al. explain why companies can’t turn customer insights into growth.

— François Maillet of explains how to use MLDB for machine learning. MLDB looks like an exciting project.

— Emmanuelle Rieuf reviews Cathy O’Neil’s Weapons of Math Destruction. So does Jo Craven McGinty in the Wall Street Journal.

Microsoft’s Big Bet on FPGAs

— Top analysts chew over Microsoft’s announcement that it uses Field Programmable Gate Arrays (FPGAs) to accelerate servers in its data centers. Karl Freund of Moor Insights and Strategy dissects Microsoft’s approach. In The Next Platform, Stacey Higginbotham delivers a tick-tock covering how MSFT decided to place its bet.

Methods and Techniques

— Alex Handy lists a collection of resources in ML, DL, and AI.

— A community of contributors offers an excellent open guide to Amazon Web Services, including Amazon Machine Learning.

Health and Medical Applications

— Jennifer Bresnick explains the potential impact of Blockchain, IoT and ML on healthcare.

— HealthNextGen, a startup that specializes in ML for health care, announces a partnership with Charité – Universitätsmedizin in Berlin, Europe’s largest university hospital.

— The National Institutes of Health awards a grant of $1.2 million to Xi Luo of Brown University and colleagues at Johns Hopkins and Yale. The grant funds ML-driven research into brain scans and brain function.

Software and Services

— Serdar Yegulalp reports on progress towards a version of TensorFlow that runs on Windows. I wonder if it will force you to upgrade to Windows 10.

— Bernd Bischl et. al. describe mlr, a machine learning framework for R with more than 160 learners and support for parallel high-performance computing.


— Nielsen adds a machine learning capability to the Nielsen Marketing Cloud.

— Wall Street mulls Tesla’s partnership with NVIDIA.

Roundup 10/21/2016


Machine learning (ML) and deep learning (DL) content from the past 24 hours. Plus, some AI stuff.

Patrick Thibodeau reports that according to Gartner, by 2020 you will say more to a machine than to your spouse. Some jokes write themselves.

Methods and Techniques

— Tim Spann says that if you can learn to play Atari you can learn TensorFlow. WTF is Atari?

— Fujitsu announces that it has developed an approach to deep learning with graph-structured data.


— Baidu bets big on Xilinx FPGAs for machine learning.

Autonomous Vehicles

— Tesla announces that its new cars will include all of the hardware needed for level 5 autonomy. The software isn’t available yet but will be added through over-the-air updates.

— In The Washington Post, Michael Laris interviews Costa Samaras of Carnegie Mellon University, who explains the implications of Tesla’s announcement.

— Michael Alba analyzes automotive supplier Denso’s partnership with vision processing startup THINCI.

Marketing and Ad Tech

— Santanu Kolay of Turn says that AI in Ad Tech is bullshit.

— No it isn’t, says Doug Conely, Chief Strategy Officer of Exponential. Guess what his company sells.

— Linsey Morse explains machine learning to marketers. In other words, without math.

— Outsell adds ML-driven content personalization to its spam engine.


— In Wired, Matthew Reynolds describes how ML will supercharge cyberattacks.

— Dan Richman interviews Danny Lange, Uber’s head of machine learning, who discusses the role of ML at Uber. He says they use Spark MLlib.

— TravelBank uses machine learning to predict expenses for business travel, which seems like overkill. All they need to do to estimate expenses is add up the costs of the booked services and multiple the travel days by a per diem.

— Kim Kyung-Jin describes AI applications in banks.

— Speaking of banks, in American Banker, Penny Crosman describes AI-driven advances in authentication.


— Lexalytics releases Salience 6.2, which includes enhanced NLP for sentiment analysis. The software can also analyze text with emojis.🙂


— Neurensic announces the availability of SCORE, an AI-driven platform for regulatory compliance in financial services.

— Loraine Lawson describes Sift Science’s ML-driven fraud prevention platform.

— CognitiveScale combines AI, Blockchain and Big Data for a buzzword hat trick.

Roundup 10/20/2016


Machine learning (ML) and deep learning (DL) content from the past 24 hours. At ML/DL, it’s nothing but machine learning and deep learning 24/7.

Microsoft AI is on a Roll

— Using CNTK, MSFT researchers achieve parity with humans in speech recognition; medialanche ensues.

— Last week, MSFT won first place in the COCO Image Segmentation Challenge.


— In HBR, Tom Davenport explains how to introduce AI into your organization. The next generation of AI will introduce itself.

— InsideBigData reports from the O’Reilly AI Conference, noting serious discussion about the drawbacks of deep learning and some attractive alternatives.

Methods and Techniques

— In Dataconomy, Kavitha Mariappan describes the ideal platform for ML.

— Adit Deshpande provides a two-part introduction to convolutional neural networks. Part One. Part Two.

ActiveClean, a collaborative between Columbia University’s WuLab and Berkeley’s AMPLab, offers an open source tool that cleans data for machine learning applications.

— Lior Shkiller describes how to capture semantic meanings in text with DL.

— Grant Ingersoll interviews the authors of Deep Learning: A Practitioner’s Approach, who explain the topic.

— In a podcast, Ben Lorica interviews Natalino Busa, who describes developments in feature engineering and predictive analytics.

Interesting Novelties

— Folks at Northwestern develop an algorithm that analyzes your Twitter feed and predicts whether you will vote for Trump or Clinton. (Gary Johnson and Jill Stein could not be reached for comment.) Of course, you already know how you’re going to vote.

— NYU researchers use ML to identify Yik Yak users.


— Nick Heath explains Microsoft’s bet on FPGAs in its Azure data centers.

— Cirrascale announces the future availability of new GPU-accelerated servers.

Health and Medicine

— Researchers at Colorado State use machine learning to identify invasive osteosarcoma cells.


— Stripe uses machine learning to detect and prevent fraud in its new Radar service.

— Glassdoor launches a new tool that uses machine learning to estimate a user’s fair salary. I suspect that complaints about errors will be asymmetric.

— LinkedIn enhances the endorsement feature with ML to make it more relevant.

— Chris Baker, Managing Director of SAP Concur, explains why AI will soon target expense reports. Consider yourself warned.

— John Dix interviews Suresh Acharya, head of JDA Labs, who discusses self-learning supply chains.

— RTB House, a marketing tech company that specializes in retargeting, announces a new conversion model built with deep learning.

— Dean Tang argues that AI will unlock IoT. Hopefully, something will.

— Martech startup CallRail releases Conversation Intelligence, an ML-driven service that automatically qualifies call leads.


— Citi Ventures invests in Feedzai, a startup that uses ML to fight fraud.

— Ravelin, another startup that fights fraud, lands a tiny “A” round.

« Older Entries