The University of Waterloo, Compute Canada and Compute Ontario unveiled the largest supercomputer at any Canadian university. Located at Waterloo, it will provide expanded resources for researchers across the country working on a broad range of topics, including artificial intelligence, genomics and advanced manufacturing. Named Graham, the supercomputer can handle more simultaneous computational jobs than any other academic supercomputer in Canada, ultimately generating more research results at one time. With its extraordinary computing power and a storage system of more than 50 petabytes -- or 50 million gigabytes -- Graham can support researchers who are collecting, analyzing, or sharing immense volumes of data. "Research and innovation have helped define the University of Waterloo, and will remain important priorities for our future," says Feridun Hamdullahpur, president and vice-chancellor of Waterloo.

There seems to be no escaping a future where robots and artificial intelligence (AI) get to perform more and more jobs, as sci-fi films warn. But Sen. Bam Aquino is not too excited about reaching such technotopia, as he raised concerns about AI completely doing away with human intelligence and leaving thousands of Filipino workers jobless. Aquino, who chairs the Senate committee on science and technology, filed a resolution on Friday seeking an inquiry into the government's preparedness to address the negative effect of AI on the country's workforce. The Department of Science and Technology (DOST) has been tapped to develop an AI program in the Philippines to optimize mass production, particularly in the country's manufacturing sector. Artificial intelligence is defined as the capability of machines to imitate human cognitive functions in doing simple to complex tasks.

French presidential candidate Emmanuel Macron's campaign said Friday it has been hacked -- 9GB of authentic and fake emails were posted Friday to Pastebin, a document-sharing site, from a user named EMLEAKS, according to Reuters. The campaign said this is "déjà vu" of the kind of "democratic destabilization" that took place in the U.S. last year. Why it matters: It's less than two days before the French presidential election comes to a close, and the way that works in France includes a ban on campaigning over the weekend. The hack came hours before the ban sets in. What to watch: The polls.

SINGAPORE: The National Research Foundation (NRF) will be investing up to S$150 million into a new national programme aimed at boosting Singapore's artificial intelligence (AI) capabilities over the next five years. Called AI.SG, the initiative will see Singapore-based research institutes partner AI start-ups and companies developing AI products to grow knowledge in the space, create tools and develop talent to power the country's AI efforts. This was announced by Minister for Communications and Information Yaacob Ibrahim at the opening of InnovFest Unbound event on Wednesday (May 3). "AI.SG will do three key things – first, address major challenges that affect both society and industry. Secondly, invest in deep capabilities to catch the next wave of scientific innovation. And finally, to grow AI innovation and adoption in companies - an initiative most pertinent to our business community," said Dr Yaacob.

Some of the world's wealthiest and most influential leaders came to California this week for the Milken Institute Global Conference, a wide-ranging review of issues permeating economics and politics, with topics ranging from agriculture to mortgage markets to international trade and alliances, plus a long look at what the future will hold. Of the 4,000 VIPs who attended -- invitations are highly selective, and tickets topped out as high as $50,000 -- one of the most intriguing questions under discussion was one that almost no one could readily answer: What effect will robotics and artificial intelligence have on our lives and on the world's business, and how rapidly will this next technological revolution take place? The Milken Institute Global Conference, an annual event for the past 20 years, has grown steadily into a unique gathering: individuals with the capital, power and influence to move the world forward meet face-to-face with those whose expertise and creativity are reinventing industry, philanthropy and media. This year's meeting in Beverly Hills, California, amounted to a peer review of President Donald Trump's first 100 days in office. Four members of Trump's Cabinet took part.

In the recent film "Hidden Figures," a group of women known as "computers," help mastermind NASA's effort to put a man into space and eventually on the moon. Without their hard work and brilliance, America's famed "moonshot" might never have succeeded. By the end of the film, their roles have largely been replaced by a computer that is capable of making millions of calculations per second -- a feat that far surpasses their abilities. Instead of despairing that they might find themselves without a job, these women adapted: they learned computer programming, thereby ensuring their ongoing relevance despite the growing role of computers at NASA. The advent of artificial intelligence, deep learning, and automation has everyone from Elon Musk to Stephen Hawking worried about the consequences.

A key limitation of sampling algorithms for approximate inference is that it is difficult to quantify their approximation error. Widely used sampling schemes, such as sequential importance sampling with resampling and Metropolis-Hastings, produce output samples drawn from a distribution that may be far from the target posterior distribution. This paper shows how to upper-bound the symmetric KL divergence between the output distribution of a broad class of sequential Monte Carlo (SMC) samplers and their target posterior distributions, subject to assumptions about the accuracy of a separate gold-standard sampler. The proposed method applies to samplers that combine multiple particles, multinomial resampling, and rejuvenation kernels. The experiments show the technique being used to estimate bounds on the divergence of SMC samplers for posterior inference in a Bayesian linear regression model and a Dirichlet process mixture model. This paper builds on a growing body of work begun by [1] and [2] into estimating upper bounds on KL divergences between a sampler's output distribution and the posterior. In variational inference, the KL divergence of the variational approximation is the gap between the variational lower bound and the log-evidence.

These data are often available in large quantities across U.S. states and require Big Data techniques to uncover complex hidden patterns. We design two different classes of models suitable for large-scale analysis of mortality data, a Hadoop-based ensemble of random forests trained over N-grams, and the DeepDeath, a deep classifier based on the recurrent neural network (RNN). We apply both classes to the mortality data provided by the National Center for Health Statistics and show that while both perform significantly better than the random classifier, the deep model that utilizes long short-term memory networks (LSTMs), surpasses the N-gram based models and is capable of learning the temporal aspect of the data without a need for building ad-hoc, expert-driven features. Many of the scientific discussions and studies in biomedical and healthcare domains address tasks whose end goal is to prevent death or diseases. Since the emergence of the big data science, numerous machine learning based techniques and technologies have been proposed and applied to improve human health by solving different computational challenges that we face today.

This paper introduces the probabilistic module interface, which allows encapsulation of complex probabilistic models with latent variables alongside custom stochastic approximate inference machinery, and provides a platform-agnostic abstraction barrier separating the model internals from the host probabilistic inference system. The interface can be seen as a stochastic generalization of a standard simulation and density interface for probabilistic primitives. We show that sound approximate inference algorithms can be constructed for networks of probabilistic modules, and we demonstrate that the interface can be implemented using learned stochastic inference networks and MCMC and SMC approximate inference programs.

HOUSTON – An arriving passenger uses a biometric scanner at George H. W. Bush Intercontinental Airport February 1, 2008 in Houston, Texas. Under President Donald Trump, technology companies have started cashing in on a little-noticed government push to ramp up the use of biometric tools -- such as fingerprinting and iris scanners -- to track people who enter and exit the country. Silicon Valley firms that specialize in data collection are taking advantage of a provision tucked into Mr. Trump's executive order on immigration, which included his controversial travel ban, that called for the completion of a "Biometric Entry-Exit Tracking System" for screening travelers entering and leaving the United States. The tracking system was mandated in a 1996 immigration law passed by Congress but never fully implemented by Trump's past three predecessors. In Trump's first months in office, federal courts blocked the sections of his original and revised immigration orders that called for a temporary travel ban on visitors from seven majority Muslim countries.