stock quote board

AI & Machine Learning News. 03, September 2018

Baidu launches EZDL, an AI model training platform that requires no coding experience

Without the technical know-how and the right tools, training machine learning algorithms can be an exercise in frustration. Luckily, for folks who don’t have the wherewithal to wade through the jargon, Baidu this week launched an online tool in beta — EZDL — that makes it easy for virtually anyone to build, design, and deploy artificial intelligence (AI) models without writing a single line of code.

Baidu’s EZDL was built with performance, ease of use, and security in mind, said Youping Yu, general manager of Baidu’s AI ecosystem division, and it targets three broad categories of machine learning: image classification, object detection, and sound classification. It’s aimed at small and medium-sized businesses, with the goal of “breaking down the barrier” to allow everyone to access AI “in the most convenient and equitable way,” Yu said.
2018-09-01 00:00:00 Read the full story.

CloudQuant thoughts… “Super Simplification”. By limiting the roles to three well-established ML categories and steering the users upload process (20-100 labeled images, or more than 50 labeled audio files), Baidu has “Super Simplified” machine learning for businesses. We can all identify some simple ML processes we could design with such a basic but simple to use system. “Generated algorithms can be deployed in the cloud and accessed via an API, or downloaded in the form of a software development kit that supports iOS, Android, and other operating systems.”

Super Simplification is the way of the future. Developing tools that make life easier for the end user to achieve massive results in a short space of time is always the ideal software goal. The type of things ordinary users can achieve with Image Software like Photoshop or Music Software like GarageBand was unimaginable twenty years ago. We are aiming to bring some of this simplification to the Auto-Trading model generating world. We believe we have hit the ideal point between making the data easy to use without making it so basic that your creativity is limited. Watch this space!

 

Detecting ‘deepfake’ videos in the blink of an eye using Machine Learning

A new form of misinformation is poised to spread through online communities as the 2018 midterm election campaigns heat up. Called “deepfakes” after the pseudonymous online account that popularized the technique – which may have chosen its name because the process uses a technical method called “deep learning” – these fake videos look very realistic.
2018-08-30 Read the full story (CNBC).
2018-08-30 Read the full story (ibtimes).

How Machine Learning Developed the Face of MCU’s Thanos

The very look of Thanos has changed slightly from movie to movie. One of several visual effects companies working on Infinity War, Digital Domain (which worked on the live-action Beauty and the Beast) wanted to include more of Josh Brolin’s facial features in Thanos’ character. Marvel Studios started working with the company several months before the official shooting for effects testing. That’s where AI stepped in to help with the “movie magic.”

Masquerade, a new machine learning software, was used to obtain the desired effect. Between 100 and 150 tracking dots were attached to the actor’s face and then recorded by a pair of HD cameras. These are actually pretty low-quality recordings, but they serve their purpose. The recordings are then sent through the machine learning algorithm which contains a vast collection of high-resolution facial scans featuring a broad range of emotions.
2018-09-02 16:55:48.997000+00:00 Read the full story.

 

CloudQuant Thoughts… Unfortunately AI and ML have already made it so that we can no longer trust what we see. Always be skeptical, always check your sources. The tech seen above, from DeepFake to Fake dancing may look obviously fake right now but the technology used to create the new Thanos at an extreme HD cinema resolution will be on our desktops or even our phones before you know it.

 

Predicting Popularity of The New York Times Comments in R (Part 1)

The New York Times (NYT) has a large reader base and plays an important role in shaping public opinion and outlook on current affairs and also in setting the tone of the public discourse, especially in the U.S. The comments sections for articles in the NYT are quite active and give insights to readers’ opinions on the subject matter of the articles. Each comment can receive other readers’ recommendations in the form of upvotes. Challenges for NYT moderators :

  • Up to 700 comments per article with NYT moderators manually reviewing ~12,000 comments in a day.
  • Moderators need to make faster decisions on screening and sorting comments based on their predicted relevance and popularity.
  • Finding an easier way to group similar comments and maintain a useful conversation among readers.

2018-09-04 00:55:55.107000+00:00 Read the full story.

CloudQuant Thoughts… An interesting article by Sakshi Gupta, in R for her capstone project for her big data certification (Ryerson University, Toronto). Of particular interest was the use of a BING word sentiment scoring system. Perhaps you can adapt this to scan news articles coming through with our trading data at app.cloudquant.com, or perhaps you can just use our pre-classified sentiment data from various high-quality sources.

 

5 Tech Hardware Stocks to Outperform: MarketWatch

While the market is swooning over hot tech names in new markets like artificial intelligence (AI), cloud computing, autonomous driving and virtual reality, one less-exciting segment of the tech space is set for major gains, according to a recent story by MarketWatch that highlighted five hardware stocks to buy in 2018. (See also: Chip Stocks on Verge of Big Breakout: Todd Gordon.)

While the big-picture trends make emerging technology markets compelling, MarketWatch’s Jeff Reeves highlighted the “rather boring but equally powerful subsector of high-tech hardware,” which has had its fare share of outperforming companies in recent years. Within the group, comprised of companies that rely on the sale of chips, semiconductors and related components, he likes Advanced Micro Devices Inc. (AMD), Micron Technology Inc. (MU), Seagate Technology PLC (STX), NVIDIA Corp. (NVDA) and Pure Storage Inc. (PSTG).
2018-08-30 12:15:00-06:00 Read the full story.

CloudQuant Thoughts… Do you agree with Jeff? A number of these symbols have already doubled or tripled in the last two years. I would say they have already achieved “major gains”. Can you do better, can you come up with a list algorithmically? Give it a try on our free backtesting system (Python knowledge required) at app.cloudquant.com.

 

Facebook is using unsupervised machine learning for translations

Facebook has begun using unsupervised machine learning to translate content on its platform when it doesn’t have many examples of translations from one language to another — such as from English to Urdu. The method was devised by Facebook AI Research (FAIR) and is being used on the platform in a collaborative effort between FAIR and the Applied Machine Learning division of the company, FAIR Paris lab director Antoine Bordes told VentureBeat in a phone interview.

The approach performs about as well as supervised models with 100,000 translations from one language to another, and it outperforms systems for language pairings for which Facebook has few examples. “When you are on cases like English-Urdu, where there’s very few [translations], there we show that our system is better than the supervised system. So it’s better to train an unsupervised system than a supervised system that doesn’t have enough data,” Bordes said.
2018-08-31 00:00:00 Read the full story.

Python “Speech Recognition Data Collection” with Youtube API, FFMPEG, Mask-RCNN and Google Vision API

With the rapid development of Machine Learning, especially Deep Learning, Speech Recognition has been improved significantly. Such technology relies on a large amount of high-quality data. However, models built for less-popular languages perform worse than those for the popular ones such as English. This is due to the fact that there is only a limited training dataset, and that it is hard to collect high-quality data effectively.

While Mozilla has launched an open-source project called Common Voice, which encouraged people to contribute their voices last year, most people are either not aware of this project, or not willing to participate in it.  Thanks to the abundance of TV shows and dramas available on Youtube, it is possible to collect speech recognition data in a highly efficient manner with almost human involvement. This blog post will show you how to efficiently collect speech recognition data for any language.
2018-08-30 Read the full story.

CloudQuant Thoughts… Lots of people are trying to solve the translation to/from less-popular languages. Facebook is relying on an unsupervised system to carry out these translations, and very successful it sounds with Antoine Bordes saying “We could go now on a planet where people speak a language that nobody else speaks — okay, the aliens — and you can actually go and try to have a decent translation of what is said there”.  黃功詳 Steeve Huang’s process is to go through a large number of individual steps to “create a data set” for training a speech recognition system using TV shows with subtitles on YouTube. He downloads them, removes the audio via FFMPEG, then removes individual frames and processes them for characters that have been encoded as subtitles on the video. Finally, he produces an audio file and a subtitle file that can be passed to a Speech Recognition training system. It is a meticulously planned process that seems to work extremely well for him. Major Kudos!

 


 

AiServe is developing assistive AI that ‘learns to walk like a human’

Visual impairment is one of the world’s most common ailments. An estimated 253 million people live with vision problems, with blindness affecting 36 million of those. Canes, guide animals, and specially adapted crosswalks make navigating busy sidewalks and city streets a little easier, but they’re not always practical.

The folks behind Berlin-based AiServe believe that artificial intelligence — specifically, natural language processing and computer vision — might be able to lend an unobtrusive helping hand. The system will run on a wearable with a camera, microphone, and a battery that lasts a few hours on a charge. As it ingests new visual data, it’ll start to recognize sidewalks, corners, and pathways with greater confidence, and in time it will map out entire city blocks and neighborhoods. The computer vision algorithm’s data will inform a navigational component that, through voice commands and other cues, will help wearers get from one place to another. Its instructions will be much more precise than most mapping apps, Madico said — rather than naming particular streets or thoroughfares, it’ll say something like “Make a left turn at this corner” or “Walk straight ahead for 100 feet.”
2018-09-01 00:00:00 Read the full story.

 

5 Cloud Computing Trends To Prepare For in 2018 – Hacker Noon

Cloud computing can literally be anything that allows you to achieve development tasks or run software through some other service provider over the internet. Cloud computing can help shorten development times, use less human resources, and provide service and availability guarantees to your clients. We had a chat with Justin Kilimnik, Director, Technology Advisory at EY to see which trends he’s excited about this year. “What key trends in cloud computing do you see emerging in the year ahead?”

  1. Docker
  2. Micro-service architecture and solutions
  3. Function-as-a-Service & Serverless
  4. Machine Learning Optimised platforms, IoT architectures
  5. Poly-cloud strategies

2018-09-03 16:53:50.504000+00:00 Read the full story.

 

AdobeStock_15670237510 Big Financial Technology Trends for 2018

2018 promises to be the year we see the culmination of some key technologies — from blockchain and intelligent AI, to design thinking and the cloud. Here are the 10 biggest trends identified by reports from Synechron and Capgemini.

  1. Massive Investments in Digital Transformation
  2. The Frontiers of Innovation: AI & Blockchain
  3. Digital-Only Banks Become a Real Threat
  4. Design Thinking
  5. Real-Time Risk Decisions
  6. Alternative Lenders Leverage Alternative Data
  7. RegTech
  8. Big Data Gets Even Bigger
  9. Connecting With Third-Party Providers to Drive Customer-Centricity
  10. The Cloud: Creeping Into Every Corner

2018-09-04 Read the full story.

 

Trump, forget Google — focus on national security AI

This week, President Trump took shots at Google for what he calls unfair search results for his name and unfair treatment of conservatives by Silicon Valley liberals. In this same vein, he talked about how some people see “an antitrust situation” with Google, Amazon, and Facebook.

Before Trump’s latest Twitter tirade began, on Sunday the New York Times reported that Defense Secretary Jim Mattis wrote a memo to President Trump earlier this year asking him to create a national strategy for AI akin to the kind China has created. China’s strategy was introduced last year and aims to make China the world leader in AI by 2030, in part through “military-civil fusion” with companies like Baidu and Tencent.

But even if he does actually believe Google treated him unfairly, it may not be in the best interest of the United States to argue with a company closely associated with the growth of AI and tools like TensorFlow. Right now, he should probably be listening to his defense secretary and thinking about what a national AI strategy should look like for the United States, and exploring the topic with companies like Google. If you believe, as Vladimir Putin does, that the nation that leads in AI will control the world, apparently there’s a lot at stake, and national security is a president’s first responsibility.
2018-08-31 00:00:00 Read the full story.

 

Thinking differently about data: the new opportunity for banks

Often described as the ‘new oil’ or even the currency of the digital economy, data is key to the business strategy of every organization. For banks, building an understanding of how to harness the variety, velocity and volumes of data available to them will dictate the difference between success and failure in the very near future.

To understand the forces at work here, it helps to paint a picture of why and how data is front and center of business planning. One reason is the absolute magnitude of data available: it’s said that 90% of all data that exists today was generated in the last two years. Data is created by social media, in the cloud, by IoT devices, on increasingly open and hyper-connected IT systems and is accessible via high performance network bandwidths.
2018-09-03 00:00:00 Read the full story.

 

High Hopes for Artificial Intelligence Don’t Always Match Reality

If you’re worried about your Echo Dot becoming self-aware and morphing into a job-stealing killing machine, worry not: That’s not happening anytime soon.

From Amazon’s Alexa to Apple’s Siri, tech companies love to talk up the limitless potential of AI and machine learning to solve all manner of problems, great and small. Salesforce’s Einstein tool, described as an AI layer to its core product suite, offers suggestions to speed up sending emails and other business processes. Facebook CEO Mark Zuckerberg has repeatedly suggested that AI will solve the social network’s many issues with data security and abuse. Meanwhile, Alphabet’s DeepMind unit – the result of a $500 million acquisition in 2018 – has gotten really, really good at board games.

But the line between technology, marketing-speak and overactive imaginations isn’t always that clear. AI has the potential to excite investors — some have even speculated that AI businesses could someday be worth five to ten times more than today’s consumer Internet companies. There are serious roadblocks to getting there, and one is a shortage of AI expertise.
2018-08-31 20:59:03-04:00 Read the full story.

 

An army of bots supporting Sweden Democrats is growing explosively ahead of September’s election

The Swedish Defence Research Agency FOI issues a warning with less than two weeks until the general election. The number of fake Twitter accounts discussing Swedish politics is soaring – and almost every other is tweeting support for the Sweden Democrats.

  • Fake accounts tweeting about Swedish politics have doubled in number ahead of the country’s general election.
  • A much larger than proportionate share is expressing support for the right-wing populist party Sweden Democrats.
  • This according to a new study by the Swedish Defence Research Agency FOI.

2018-08-29 16:54:48 Read the full story.

 

LG To Strengthen AI, Robotics Business Amid Struggling Smartphone Unit

LG CEO Jo Seoung-Jin showed up at Internationale Funkausstellung (IFA) Berlin late last week and told reporters about LG’s plans moving forward. According to him, the company is strengthening its AI and robotics business this year by increasing the number of its engineers and expanding the support base for its AI technology and robots. “The world is heading toward an era of AI and that embracing the new trend is critical”.
2018-09-03 09:22:12-04:00 Read the full story.

 

JPMorgan Hires Top AI Exec Away From Google

As the race on Wall Street to deploy artificial intelligence (AI) ramps up, JPMorgan Chase & Co. has shown how serious it is about its next-gen tech initiative with another major hire this week. The bank hired Apoorv Saxena, Alphabet Inc.’s head of product management for cloud-based AI, according to a memo obtained by CNBC. The senior Google executive will start at JPMorgan on Aug. 31 as head of AI and machine learning services and will also be responsible for leading the firm’s AI-powered asset and wealth management technology initiative. Recruiting talent has been an integral part of a larger strategy among traditional financial institutions in order to develop AI for improved and automated services like fraud detection, internal operations and loan approval.
2018-08-29 11:41:00-06:00 Read the full story.

 

Lite Intro into Reinforcement Learning

This is a brief introduction into Reinforcement Learning (RL) going through the basics in simplified terms. We start with a brief overview of RL and then get into some practical examples of techniques solving RL problems. In the end you may even think of places you can apply these techniques. I think we can all agree building our own Artificial Intelligence (AI) and having a robot do chores for us is cool… so let’s get to it!

2018-09-03 21:10:09.394000+00:00 Read the full story.

 

Is hybrid cloud the best of both worlds? Five experts explain how to get it right

Fervent cloud computing evangelists used to look down their noses at the notion of hybrid cloud, seeing the approach a cop-out for slow-moving and shallow-minded tech organizations. Things have most certainly changed. We invited five experts to help attendees at the 2018 GeekWire Cloud Tech Summit understand how big companies and growing startups are implementing hybrid cloud strategies.

  • Alex Legault, Associate Director of Products, PitchBook
  • Jin Zhang, Director, Product Management VMware & Hybrid Computing, Amazon
  • Nicholas Criss, Sr. Manager, Cloud Center of Excellence, T-Mobile
  • Madhura Maskasky, Co-founder, Platform9
  • Anthony Skinner, CTO at iSpot.tv

2018-09-02 19:37:33-07:00 Read the full story.

 

Apple’s Pious Privacy Pledges Ring Hollow

Privacy is the new battleground. Sounds simple, but it’s not. This war is really about the power of networks, and legacy versus disruption. There is no debate. Apple is succeeding in China. Oddly, its advocacy for user data privacy does not extend to the People’s Republic. Apple transferred the operation, including the security keys, of its Chinese iCloud service to Guizhou-Cloud Big Data, in February. GCBD is a state-owned enterprise.
2018-09-03 08:00:00-04:00 Read the full story.

 

Musk Personally More Likely Than Tesla to Have to Pay Back the Shorts

That might have been a billion-dollar tweet. If anyone is going to pay for Elon Musk’s misleading tweet that he had secured the funding to take Tesla (TSLA – Get Report) private, it will likely be the Tesla CEO himself, and not his electric car company, lawyers said.

Short sellers who target Tesla because of the view that the company’s stock price isn’t justified by its profit prospects may have lost about $1.3 billion on August 7, the day of the misleading tweet, although most investors with short positions in Tesla didn’t cover their positions that day, according to financial data analytics firm S3 Partners. Total losses in August for the short sellers may be have been as much as $3 billion, S3 said. The losses prompted at least two lawsuits to be filed in federal court in San Francisco, and law firm Pomerantz LLP has even provided a draft complaint on its website for investors contemplating suing both Musk and Tesla.
2018-08-28 16:45:10-04:00 Read the full story.

 

Google releases AI-powered Content Safety API to identify more child abuse images

Google has today announced new artificial intelligence (AI) technology designed to help identify online child sexual abuse material (CSAM) and reduce human reviewers’ exposure to the content. The move comes as the internet giant faces growing heat over its role in helping offenders spread CSAM across the web. Last week, U.K. Foreign Secretary Jeremy Hunt took to Twitter to criticize Google over its plans to re-enter China with a censored search engine when it reportedly won’t help remove child abuse content elsewhere in the world.

News emerged last year that London’s Metropolitan Police was working on a AI solution that would teach machines how to grade the severity of disturbing images. This is designed to solve two problems — it will help expedite the rate at which CSAM is identified on the internet, but it will also alleviate psychological trauma suffered by officers manually trawling through the images. Google’s new tool should assist in this broader push.
2018-09-03 00:00:00 Read the full story.

 

Silicon Valley is under pressure again, this time over online paedophilia

Home Secretary Sajid Javid has warned about the sheer volume of pedophilic content available online. He is expected to announce new measures to combat the problem later on Monday. The UK’s National Crime Agency revealed that it received over 80,000 industry referrals for child sex abuse images in 2017, which represents a 700% increase since 2012. The agency asked tech companies to cooperate with law enforcement more to combat online child sexual abuse. Google also announced on Monday that it is rolling out a new AI tool to help NGOs and industry partners track down child abusers.
2018-09-03 00:00:00 Read the full story.

 

Google launches AI tool to identify online child sex abuse images (Paywall)

Google has unveiled new technology designed to identify images of child abuse, as the Silicon Valley giant sought to stave off mounting pressure for action to tackle exploitative content. The US company said it was releasing free ‘cutting edge’ artificial intelligence software that would help web moderators root out abusive content on the Internet on a large scale. Google said the technology “significantly advances our existing technologies to dramatically improve how service providers, NGOs, and other technology companies review this content at scale.”
2018-09-03 00:00:00 Read the full story.

 

Crowdsourcing in the age of artificial intelligence: How the crowd will train machines

It was over 10 years ago that I was introduced to the concept of crowdsourcing. I was a student at London Business School when a professor one day came into the classroom with a jar of pennies. He asked us each to take a look at the jar and guess the correct amount of money inside. The jar went around the classroom and I gave it an estimate of £30 in good faith. The professor duly wrote down each of our 100 guesses on the whiteboard and then opened a sealed envelope where the real amount was revealed: £18.76.

While my initial lesson learned was that I shouldn’t ever try a career in penny guessing, the amazing surprise was still in store for me: The professor calculated the average of all our 100 guesses and it magically came down to £18.76. The wisdom of the crowd was spot on and was better than 99 percent of our own estimates (only one of us actually guessed the right amount).
2018-09-01 00:00:00 Read the full story.

 

The man in the red polo: Meet Scott Guthrie, Microsoft CEO Satya Nadella’s front-line general in the cloud wars with Amazon and Google

Meet Scott Guthrie, the executive VP of Microsoft’s cloud and artificial intelligence business. He’s a long-time Microsoft exec, and a trusted lieutenant to CEO Satya Nadella — the two worked together to get the Microsoft Azure cloud business off the ground. He’s as well known for his red polo shirt as he is for his technical acumen and leadership skills.

In this interview, Guthrie discusses what he learned from working with Netflix on streaming, how he works with Nadella, and what he believes is Microsoft’s secret weapon in the cloud wars with Amazon and Google.
2018-09-03 00:00:00 Read the full story.

 

Weekly Selection — Aug 31, 2018 – Towards Data Science

 

  • How to construct valuable data science projects in the real world
  • Convolutional Neural Networks: The Biologically-Inspired Model
  • Making Music: When Simple Probabilities Outperform Deep Learning
  • How to Run Parallel Data Analysis in Python using Dask Dataframes
  • Named Entity Recognition and Classification with Scikit-Learn
  • A Day in the Life of a Marketing Analytics Professional
  • Changing The Engineer’s Mindset : From How to Why
  • Tutorial: Double Deep Q-Learning with Dueling Network Architecture
  • Automatic Speech Recognition Data Collection with Youtube V3 API, Mask-RCNN and Google Vision API
  • How to build a non-geographical map #1

2018-08-31 13:17:44.165000+00:00 Read the full story.

 


Behind a Paywall…

Britain faces an AI brain drain as Silicon Valley raids its top universities for talent

Around a third of leading machine learning and AI specialists who have left the UK’s top institutions are currently working at Silicon Valley tech firms. More than a tenth have moved to North American universities and nearly a tenth are currently working for other smaller US companies. Meanwhile just one in seven have joined British start-ups.
2018-09-02 Read the full story.

 

Tech giants keep British artificial intelligence in mind

For mere mortals job hunting can be a slog. Updating your CV, writing a cover letter, having a phone interview, an in-person interview, a second interview. It can seem like a never-ending process. But for specialists in the red-hot field of artificial intelligence (AI), job offers rain down with little effort.

With billions of dollars flowing into the sector, around the world big technology companies are engaged in a scramble for talent.
2018-09-02 00:00:00 Read the full story.

 


This news clip post is produced algorithmically based upon CloudQuant’s list of sites and focus items we find interesting. If you would like to add your blog or website to our search crawler, please email customer_success@cloudquant.com. We welcome all contributors.

This news clip and any CloudQuant comment is for information and illustrative purposes only. It is not, and should not be regarded as “investment advice” or as a “recommendation” regarding a course of action. This information is provided with the understanding that CloudQuant is not acting in a fiduciary or advisory capacity under any contract with you, or any applicable law or regulation. You are responsible to make your own independent decision with respect to any course of action based on the content of this post.