10 Things Everyone Should Know About Machine Learning

By Daniel Tunkelang, who led machine learning projects at Endeca, Google, LinkedIn. Originally published on Quora.

As someone who often finds himself explaining machine learning to non-experts, I offer the following list as a public service announcement.

  • Machine learning means learning from data; AI is a buzzword. Machine learning lives up to the hype: there are an incredible number of problems that you can solve by providing the right training data to the right learning algorithms. Call it AI if that helps you sell it, but know that AI, at least as used outside of academia, is often a buzzword that can mean whatever people want it to mean.
  • Machine learning is about data and algorithms, but mostly data. There’s a lot of excitement about advances in machine learning algorithms, and particularly about deep learning. But data is the key ingredient that makes machine learning possible. You can have machine learning without sophisticated algorithms, but not without good data.
  • Unless you have a lot of data, you should stick to simple models.Machine learning trains a model from patterns in your data, exploring a space of possible models defined by parameters. If your parameter space is too big, you’ll overfit to your training data and train a model that doesn’t generalize beyond it. A detailed explanation requires more math, but as a rule you should keep your models as simple as possible.
  • Machine learning can only be as good as the data you use to train it. The phrase “garbage in, garbage out” predates machine learning, but it aptly characterizes a key limitation of machine learning. Machine learning can only discover patterns that are present in your training data. For supervised machine learning tasks like classification, you’ll need a robust collection of correctly labeled, richly featured training data.
  • Machine learning only works if your training data is representative. Just as a fund prospectus warns that “past performance is no guarantee of future results”, machine learning should warn that it’s only guaranteed to work for data generated by the same distribution that generated its training data. Be vigilant of skews between training data and production data, and retrain your models frequently so they don’t become stale.
  • Most of the hard work for machine learning is data transformation.From reading the hype about new machine learning techniques, you might think that machine learning is mostly about selecting and tuning algorithms. The reality is more prosaic: most of your time and effort goes into data cleansing and feature engineering — that is, transforming raw featuresinto features that better represent the signal in your data.
  • Deep learning is a revolutionary advance, but it isn’t a magic bullet.Deep learning has earned its hype by delivering advances across a broad range of machine learning application areas. Moreover, deep learning automates some of the work traditionally performed through feature engineering, especially for image and video data. But deep learning isn’t a silver bullet. You can’t just use it out of the box, and you’ll still need to invest significant effort in data cleansing and transformation.
  • Machine learning systems are highly vulnerable to operator error. With apologies to the NRA, “Machine learning algorithms don’t kill people; people kill people.” When machine learning systems fail, it’s rarely because of problems with the machine learning algorithm. More likely, you’ve introduced human error into the training data, creating bias or some other systematic error. Always be skeptical, and approach machine learning with the discipline you apply to software engineering.
  • Machine learning can inadvertently create a self-fulfilling prophecy. In many applications of machine learning, the decisions you make today affect the training data you collect tomorrow. Once your machine learning system embeds biases into its model, it can continue generating new training data that reinforces those biases. And some biases can ruin people’s lives. Be responsible: don’t create self-fulfilling prophecies.
  • AI is not going to become self-aware, rise up, and destroy humanity. A surprising number of people (cough) seem to be getting their ideas about artificial intelligence from science fiction movies. We should be inspired by science fiction, but not so credulous that we mistake it for reality. There are enough real and present dangers to worry about, from consciously evil human beings to unconsciously biased machine learning models. So you can stop worrying about SkyNet and “superintelligence”.

There’s far more to machine learning than I can explain in a top-10 list. But hopefully this serves as a useful introduction for non-experts.

By Daniel Tunkelang, who led machine learning projects at Endeca, Google, LinkedIn. Originally published on Quora.

For more trending tech answers from Quora, visit HackerNoon.com/quora.

This content was originally published here.

NVIDIA Creates First Artificial Intelligence Rendered Virtual World – Geekologie

This is a brief video tour of an interactive virtual world that was rendered entirely by an artificial intelligence system after it was trained using a bunch of real world footage of cars driving around various cities. It’s pretty…

This content was originally published here.

MSOE announces construction of $34M artificial intelligence facility

Milwaukee School of Engineering officials on Monday announced the construction of a new, $34 million academic facility which will specialize in artificial intelligence and computational science education.

This content was originally published here.

Kevin Scott of Microsoft Hopes Artificial Intelligence Will Help His Hometown – WSJ

Microsoft’s chief technology officer, Kevin Scott, is humming Mozart’s “Sonata facile,” a melody that he recalls hearing in “Looney Tunes” cartoons. A computer could easily learn to play it, he says, but such a rendition probably wouldn’t elicit an emotional response from a listener. “It’s not about the notes; it’s about this connection…with the audience,” he says. “I don’t know if that’s going to be possible with a machine.”

Mr. Scott, 48, has been working on artificial intelligence for most of his working life. He is acutely…

This content was originally published here.

Masa Son’s tech empire is being rocked by the pandemic. Don’t count him out just yet – CNN

For Son, it is a surprising change of tack.
Startled by the company’s falling stock price earlier this month, SoftBank managers decided something drastic had to be done, according to Bernstein analyst Chris Lane. Describing the shift as a “radical change in strategy” Lane upgraded his recommendation on SoftBank from buy to outperform, noting that even if the value of Vision Fund startups halve, the stock is still cheap.
Part of that significant change is SoftBank signaling a step back from Son’s high-risk style of investing. SoftBank chief financial officer Yoshimitsu Goto said on an investor call last week that the company would be very conservative with new investments given the current economic climate, according to analysts from Bernstein and CLSA that were on the call.
Rajeev Misra, CEO of SoftBank Investment Advisors, expected dozens of Vision Fund companies to IPO in the next two years.
Spokespersons from SoftBank and the Vision Fund did not respond to requests for comment.
Credit ratings agency Moody’s was a lot more pessimistic about SoftBank’s new approach than some analysts. Moody’s pushed the company’s credit further into junk status, citing its “aggressive financial strategy.”
Selling SoftBank’s “highly-valued” stakes in established companies such as China’s Alibaba (BABA) or US mobile provider Sprint (S), just as the market is getting roiled by the pandemic, would drive down the value of its portfolio, Moody’s said. SoftBank said the transactions will take place over the next year, but has not indicated what it is likely to sell.
“Asset sales will be challenging in the current financial market downturn, with valuations falling and a flight to quality,” Motoki Yanase, Moody’s vice president and senior credit officer, said last Wednesday.
SoftBank immediately pushed back on the downgrade, saying in a statement that Moody’s had “ignored” the company’s explanations. It accused the agency of having “biased and mistaken views.” Rival ratings agency S&P, by contrast, said the asset sale could be good for SoftBank’s credit quality.
The share price crash — the stock hit a near four-year low earlier this month — may have forced Son to tap the brakes on his global spending spree. But don’t count him out just yet. He’s been on the back foot before, and still stuck to his strategy of betting big on splashy tech firms and their eccentric founders.
Son lost billions when the dotcom bubble burst — 99% of his net worth was wiped out in 2000.
He also made his most famous investment the same year, putting $20 million into Alibaba, a fledgling Chinese e-commerce company. That bet turned into $60 billion when Alibaba went public in 2014. SoftBank has since sold some shares, but as of Monday, its 25.1% stake in Alibaba was still worth nearly $129 billion.
When people were criticizing internet companies 20 years ago, “I had confidence and vision in [them],” Son said at an earnings presentation in November. When it comes to artificial intelligence and the portfolio companies in the Vision Fund, “we have the same confidence in them,” he said.

SoftBank and Vision Fund investments are exposed

Still, a lot has changed since last fall. Forty percent of the Vision Fund portfolio is made up of transportation and logistics startups, companies that are among the most exposed to the global lockdowns. Son has spent some $33 billion buying stakes in companies such as Uber, Didi, Ola and Grab.
In February, at the peak of the coronavirus outbreak in China, the number of people using ride hailing app Didi plunged by nearly 60% in cities like Beijing and Shanghai, according to research company Arora Mobile. Didi’s woes could be a harbinger of what’s to come for Uber in the United States and Europe, Ola in India and Grab in Southeast Asia.
A worker wears protective clothing as he disinfects a car for Chinese ride hailing company Didi in Beijing in February.
Those ride hailing companies are facing steep losses because of the outbreak, according to Raymond Tsang, a partner with consulting firm Bain & Company. Most ride hailing firms aren’t profitable, because they’re burning through money to fend off rivals and gain market share.
With the “drain of cash” caused by the pandemic, “it will be harder for Didi, Grab, or any other mobility players to reach their breakeven point, or at least it will take them longer to sustain their economics,” Tsang said.
As of March 30, the Vision Fund was sitting on a loss of nearly $1.5 billion on its roughly 13% stake in Uber, according to CNN Business calculations. Shares in the car hailing company are down 33% from a recent high in early February, reversing a recovery Son had touted at an earnings presentation just last month.
Stakes in real estate startups, which make up 10% of the Vision Fund portfolio, will also be hit.
WeWork is particularly vulnerable as major cities where it operates shut down for weeks on end. It still has to pay long-term leases, even if businesses squeezed by the outbreak cancel contracts with the shared office space company. To make matters worse, SoftBank is now reportedly backing away from part of its $10 billion bailout of the company.


Former WeWork CEO is getting a massive payout

Former WeWork CEO is getting a massive payout 02:29

IPO exits unlikely any time soon

Earlier this month, Vision Fund head Rajeev Misra told CNBC that he expected dozens of the Vision Fund’s companies to go public in the next 18 to 24 months.
Since then, the pandemic has worsened. Investments funds should now be prepared to hold onto their startups for longer than they anticipated, according to research firm PitchBook.
Firms “will be forced to hold struggling assets and may find exit markets to be completely subdued for most assets, particularly across the strategic [merger and acquisition] and IPO avenues,” PitchBook analysts wrote in a note last week.
The coronavirus outbreak could also exacerbate a trend that began with WeWork’s botched IPO last fall.
Son admitted that he had put too much faith in founder Adam Neumann, and the balance of power appears now to have shifted “in favor of investors” and away from “founder-friendly terms” and growth at all costs, the PitchBook analysts wrote.
Push back from investors could make it harder for Son and Misra to raise the $10 billion they are reportedly seeking to support Vision Fund portfolio companies grappling with the coronavirus outbreak.
One SoftBank-backed company that needed more money has already collapsed. OneWeb filed for bankruptcy last week. The company was not part of the Vision Fund portfolio, but had received about $2 billion from SoftBank.
The internet satellite startup was in talks with SoftBank to secure new funding, but the negotiations fell apart hours before OneWeb launched a batch of satellites into orbit on Saturday, according to a Financial Times report that was confirmed by CNN Business.

Son may yet be right

Son has been saying for years that a revolution is coming, when the use of artificial intelligence fundamentally changes how people live and work.
And some are backing his vision — Chris Matthews, with CLSA, said the virus may help accelerate the global change on which Son has been fixated.
“The coronavirus is, in our view, likely to drive adoption for the new disruptive technology companies in which Softbank Group has been investing,” Matthews wrote in a note earlier this month.
Autonomously driven rides from Uber or Didi, for example, would have come in handy during this outbreak.
But Son’s desire to raise even more money to support the Vision Fund during this crisis still worries some analysts.
Atul Goyal, of Jefferies, said last week that the market clearly doesn’t think the Vision Fund has added any value to SoftBank, and if Son “remains committed to raise funds for Vision Fund, the value destruction risk [to SoftBank] remains.”
Correction: An earlier version of this story incorrectly stated that SoftBank’s OneWeb investment was held by the Vision Fund.

This content was originally published here.

Artificial Intelligence VS Machine Learning VS Data Science | Codementor

This article states the basic difference between Difference between Artificial Intelligence (AI), Machine Learning (ML) and Data Science.

This content was originally published here.

Pluralsight makes entire library of courses free for April | ZDNet

Pluralsight has announced it is giving new users free, unlimited access to its entire library of more than 7,000 technology courses for all of April.

Pluralsight co-founder and CEO Aaron Skonnard took to Twitter to make the announcement about the offering.

“I’m so excited to see what you learn, and what you create this month,” he said, encouraging everyone to stay home during a time where the world attempts to combat the novel coronavirus outbreak.

The free access to the video-based courses will only be available to those who sign up for a new account during April.

Users who sign up for a new account will not be required to provide any credit card details and will not have any watch limits.

Using their free accounts, eligible users can access course topics such as software development, IT ops, data professional, information and cybersecurity, and machine learning and artificial intelligence.

New free accounts opened after May 1, meanwhile, will only have access to a portion of Pluralsight’s library and will be required to provide payment information.

Current active subscribers, however, are not eligible for the promotion.

Companies such as Atlassian, Okta, Tableau, IBM, and Salesforce have also extended free versions of their offerings to organisations to help them during the pandemic.

Audible is also providing free services for the duration of the coronavirus pandemic through the launch of Audible Stories, a new service that provides free audiobooks for small children and teens.

“For as long as schools are closed, we’re open,” Audible said.

“Starting today, kids everywhere can instantly stream an incredible collection of stories, including titles across six different languages, that will help them continue dreaming, learning, and just being kids,” the company said.

This content was originally published here.

Salesforce Marketing Cloud brings Einstein AI capabilities to email marketers – Marketing Land

Salesforce revealed new Einstein AI capabilities available for its Marketing Cloud users, on Monday at its annual Salesforce Connections conference. The new features, available to customers now, aim to help marketers reach their customers more efficiently and effectively with email through the use of artificial intelligence.

Why we should care

Artificial intelligence is one of the hottest trends in martech right now, and it was only a matter of time before Salesforce enabled its Einstein AI in the Salesforce Marketing Cloud.

The new tools and capabilities could prove valuable to email marketers by predicting the best time and frequency to send emails. Digital marketers and organizations who embrace — and trust — the technology’s ability to make these decisions for them will need to learn to let go of making those decisions, which may be easier said than done.

More on the news

The Marketing Cloud enhancements are available as of Monday’s announcement. The new capabilities include:

Related Topics

This content was originally published here.

The latest WiFi-marketing tactic? Chatbot-enabled surveys on WiFi sign-on pages – Marketing Land

Wifi-marketing can be an effective practice for brands with physical locations, and it might be getting smarter. Bespoke, a Japanese artificial intelligence developer, announced Tuesday that the company has launched its AI-enabled WiFi chatbot in the United States. Bespoke’s technology is generally used by brands in the travel and hospitality industries.

Why we should care

Digital marketers use all kinds of tactics to engage with customers; WiFi engagement platforms like Aislelabs can be used to develop personalized experiences and deliver updates to visitors. For brands more interested in strictly collecting WiFi data, platforms like Gazelle offer solutions to deliver WiFi to customers and data to the marketers. And considering most places have WiFi, it’s an accessible means for engagement for both consumers and digital marketers.

Location-based marketing is hot right now. Artificial intelligence is heavily influencing the martech landscape. Combining the two could have a lasting effect on digital marketing. chatbots, among other LBM strategies, digital marketers who execute AI chatbots at physical locations — and do it well — will undoubtedly cause a shift in consumers’ expectations for brand interactions.

More on the news

Jennifer Videtta serves as Third Door Media’s Senior Editor, covering topics from email marketing and analytics to CRM and project management. With over a decade of organizational digital marketing experience, she has overseen digital marketing operations for NHL franchises and held roles at tech companies including Salesforce, advising enterprise marketers on maximizing their martech capabilities. Jennifer formerly organized the Inbound Marketing Summit and holds a certificate in Digital Marketing Analytics from MIT Sloan School of Management.

Related Topics

This content was originally published here.

Amazon creates voice-activated device that recognises human emotions

Amazon is developing a voice-activated wearable device that can read human emotions and tell whether people are angry or sad based just on their tone of voice. 

The wrist-worn gadget, code-named “Dylan”, will feature microphones paired with software that can detect a person’s mental state based on their voice alone. It could even be used to offer counselling to users on how to interact with other people.

The retail giant has been working on the wearable that would pair with a smartphone app and early beta testing has begun.

The device is being billed as a “health and wellness” gadget, according to internal documents seen by Bloomberg. The new technology could be capable of recording emotions such happiness, excitement, fear and disgust. 

The device is a collaboration between Lab126, the hardware development group that developed Amazon’s phone and the Alexa voice team.

While emotion detection was once considered in the realms of science fiction, robots using computer vision and artificial intelligence are getting increasingly adept at spotting human emotions from their faces. 

But voice detection is at the centre of Amazon’s ambitions. Its Echo smart speakers have proved popular and the company has said it wants to create more lifelike virtual assistants. 

This content was originally published here.