Originally posted here, where you can see all the graphics. There has been much in the news lately about the next wave of MT technology driven by a technology called deep learning and neural nets (DNN). I will attempt to provide a brief layman's overview about what this is, even though I am barely qualified to do this (but if Trump can run for POTUS then surely my trying to do this is less of a stretch). Please feel free to correct me if I have inadvertently made errors here. To understand deep learning and neural nets it is useful to first understand what "machine learning" is. Very succinctly stated machine learning is the "Field of study that gives computers the ability to learn without being explicitly programmed" according to Arthur Samuel. Machine learning is a sub-field of computer science that evolved from the study of pattern recognition and computational learning theory in artificial intelligence.

The less constrained model indeed performs better (judging by AIC, which drops from to 3696 to 3278). Autocorrelation of errors also is reduced overall. Now, with the improved models, let's finally get forecasting!

The AI is able to create computer-generated singalong based only on a festive digital photo. Dubbed'neural karaoke', the software writes its own simple composition and also provides on-screen lyrics for any budding singers to join in Can MATHS help you win at roulette? Expert reveals the... Is THIS the future of shopping? You can now play Pac-Man in Messenger: Facebook adds... Are YOU smart enough to answer Google's tricky interview... Can MATHS help you win at roulette? Expert reveals the... Is THIS the future of shopping?

TV series "Humans" shows robots taking over simple human tasks Photo: Channel 4 Over the last decade, the types and definitions of Artificial Intelligence (AI) have ranged across a wide spectrum. A future of smart homes and smart cars, driven by AI, is no longer a distant reality. This future poses questions to businesses; principal among them is how businesses can adapt to AI. The practical impact on companies, and their workforces, is an increasingly pressing issue around the boardroom. The International Federation of Robots (yes, a real thing, not a Star Wars coalition) reported last year that robot sales in 2014 increased by 29% – the biggest increase ever recorded within a year, and this trend is expected to continue.

Next time you stop for gas at a self-serve pump, say hello to the robot in front of you. Its life story can tell you a lot about the robot economy roaring toward us like an EF5 tornado on the prairie. Yeah, your automated gas pump killed a lot of jobs over the years, but its biography might give you hope that the coming wave of automation driven by artificial intelligence (AI) will turn out better for almost all of us than a lot of people seem to think. The first crude version of an automated gas-delivering robot appeared in 1964 at a station in Westminster, Colorado. Short Stop convenience store owner John Roscoe bought an electric box that let a clerk inside activate any of the pumps outside. Self-serve pumps didn't catch on until the 1970s, when pump-makers added automation that let customers pay at the pump, and over the next 30 years, stations across the nation installed these task-specific robots and fired attendants. By the 2000s, the gas attendant job had all but disappeared.

International Business Machines launched Watson solutions for enterprise in 2014. Even though close to 63% of its value stems from software (middleware and operating system) and technology services business, the fast changing delivery mechanism for software (Cloud Software as a Service) and its deployment (Infrastructure as a Service) has affected top line growth in more than one way. While IBM's middleware is loosing its sheen as application development moves to cloud, IaaS across its technology services is gaining momentum. However, IBM has carefully maneuvered itself into the next wave of tech growth that relies heavily on big data analytics with its cognitive computing system, Watson. Recently, IBM has introduced its first global consulting practice for the internet of things (IoT) on Watson.

Many products or published articles based on data science are heavily regulated, and illegal to perform or publish or sell without a special license, especially in US. You may be doing research and development on a topic considered as classified by the US government. Steganography (the art and science of hiding secret messages in images) Factoring the product of two large primes (cryptography application) Algorithms to reverse-engineer some systems (e.g.to check if a credit card number is valid) Distribution of encoding algorithms that can be used and customized by anyone, even offline Selling or designing home-made, backdoor-free, weapon-grade encryption products abroad or even in US Manufacturing and selling weapon-grade random numbers, for instance based on the digits of some transcendental numbers, using very fast algorithms (have you ever wondered why Excel random numbers were so poor - maybe they are very poor on purpose, not because those who designed them are ignorant) Algorithms to reverse-engineer some systems (e.g.to check if a credit card number is valid) Depending on how strong your algorithms are, it could be classified material, and you might not be aware of it until you get a letter from the NSA (after all, if you don't have access to classified information, there's no way for you to know whether what you do is classified or not). Designing algorithms, trained on data, to beat the IRS, using legal techniques only Publishing a list of prices for medical procedures (with price range), broken down per hospital, based on data collected online from patients, with their consent Gambling or bets that involve guessing econometric indicators, such as stock prices or indices. While this is different from a lottery in the sense that winning is based on good data science rather than pure luck (lottery is another business that is great for a data scientist, although universally illegal), it might still be illegal depending on your jurisdiction.

Do you need a chat with a Data Doctor? The Data landscape is vast and ever changing, it can also be a place where we are bombarded with buzzwords, new trends and an onslaught of tools and services. Your body needs a check-up every now and then, so does your data journey need a check-up too? The Data Lab are running drop-in sessions with industry experts, "Doctors", who offer practical, executable advice to attendees covering 3 of the current big themes in data. Each Doctor has nine 12-minute slots available during the day, so request your 1-2-1 slot quickly before they are fully booked out.

For basketball fans of a particular vintage, it conjures memories of a pint-sized shoot-first point guard who reportedly showed up for press avails under the influence of at least one mind-altering substance. Allen Iverson even established some general pop-culture visibility with his on- and off-court exploits. He was a quintessential "good news-bad news" basketball story. Though the former Philadelphia 76ers superstar is now an NBA Hall of Famer, nowadays AI is all about artificial intelligence. But this AI is more than a simple "good news-bad news" dichotomy.

Five findings from YouGov's European mega-survey (YouGov) Some more findings from YouGov's European mega-survey On that NYT democracy chart… Corrective 1 (Tom van der Meer) and Corrective 2 (Cas Mudde) Six charts that are key to understanding Italy's referendum (FT Data) Economists are prone to fads, and the latest is machine learning(The Economist) How to better communicate election forecasts -- in one simple chart (Monkey Cage, Washington Post) Clinton would have won if the United States looked like this(Kevin Hayes Wilson) Six charts that are key to understanding Italy's referendum (FT Data) How would you describe Fidel Castro? 'Warning Signs Are Flashing Red'(New York Times) Shift Your Point of View to When America Was "Better" (Flowing Data) How would you describe Fidel Castro? How would you describe Fidel Castro? 'Warning Signs Are Flashing Red'(New York Times) Shift Your Point of View to When America Was "Better" (Flowing Data) The OGN steering committee writes to Ben Gummer on the future of open government (Open Government Network) Where we've got to; the OGP Summit; & what's next (Open Government Network) Open Government Change We Can Believe In (Tom Cochran) Where we've got to; the OGP Summit; & what's next (Open Government Network) Why GOV.UK Verify is safer (Government Digital Service) Troubled Families and the Digital Economy Bill (Data in government) Why GOV.UK Verify is safer (Government Digital Service) Friday lunchtime lecture: Creating a web of cats (ODI) Christmas 2015 Was Filled With Hoverboards -- And Hoverboard Injuries (FiveThirtyEight) A Normal Distribution (of Legs): Etiquette and equality may be openly encroached by manspreading, but efficiency is the act's hidden victim (Walker Harrison) A Normal Distribution (of Legs): Etiquette and equality may be openly encroached by manspreading, but efficiency is the act's hidden victim (Walker Harrison) Would a different style of voting have changed the 2016 election? Would a different style of voting have changed the 2016 election?