Artificial intelligence (AI)–assisted electrophysiology (EP) shows promise, but even its most ardent advocates aren't ready for full-fledged endorsement--yet. A robot revolution is coming, predicted the Huffington Post in November. Citing Bureau of Labor Statistics data and an analysis by Ball State University's Center for Business and Economic Research (CBER, online June 19, 2017), the article warned that nearly half of American jobs are "vulnerable to automation." Many came couched among counsel to prepare for a changing job market, while others have joined the debate about what machines should and shouldn't be tasked with doing. Medicine hasn't been immune to the debate, leading some clinicians to worry that AI--the same brain behind automation, robotics and deep learning--might someday replace them just as it's predicted to do to workers in many manufacturing jobs.

November 15, 2017 Written by: Kareem Yusuf, Ph.D As a huge Star Wars fan, I have always wanted to have a set of supernatural powers, such as the ability to sense impending attacks; influence the thoughts of others, known as the "Jedi mind trick;" and even see the future. In essence, the Force – meaning the wisdom, expertise and ability to know exactly what will happen next and what I should do in the moment. Of course, there is only one problem with all this--our current reality. Wilson predicted, the world will be run by synthesizers, people able to put together the right information at the right time and think critically about it to make important choices wisely. And, that is exactly what IBM Watson is doing today.

Ensuring endpoint security has always been a key challenge for enterprises. But whereas it was once enough to install antivirus (AV) software across a network and expect a reasonable level of endpoint protection, this is no longer the case. With the proliferation of bring your own device policies in the workplace and the wide variety of smart devices available to end users, not to mention the growth of IoT, there are more endpoints than ever, and endpoint security has never been more under threat. Get the latest from CSO by signing up for our newsletters. Various studies put the number of security breaches originating at endpoints between 70 and 95 per cent.

Ministers have agreed a technology tie-up with Emmanuel Macron's government that will see the UK unite with France in areas such as artificial intelligence and cyber security. Matt Hancock, the Culture Secretary, will this morning announce the two countries plan to work together to pool industry and academic research. Britain and France will host a conference later this year to encourage cross-Channel investment, targeting work on AI in particular. It comes as France's new government tries to shake off its reputation for protectionism and onerous employment rules that has held back start-ups in the country and seen Paris lose out to London as a tech hub. Last year more venture capital deals were signed involving French tech companies than British ones, the first time this has happened in half a decade, although the UK raises far more money in total and dominates rankings of European "unicorns", start-ups with valuations of more than $1bn (£720m).

In this paper, we study regression problems over a separable Hilbert space with the square loss, covering non-parametric regression over a reproducing kernel Hilbert space. We investigate a class of spectral-regularized algorithms, including ridge regression, principal component analysis, and gradient methods. We prove optimal, high-probability convergence results in terms of variants of norms for the studied algorithms, considering a capacity assumption on the hypothesis space and a general source condition on the target function. Consequently, we obtain almost sure convergence results with optimal rates. Our results improve and generalize previous results, filling a theoretical gap for the non-attainable cases.

This article develops a statistical test for the null hypothesis of strict stationarity of a discrete time stochastic process. When the null hypothesis is true, the second order cumulant spectrum is zero at all the discrete Fourier frequency pairs present in the principal domain of the cumulant spectrum. The test uses a frame (window) averaged sample estimate of the second order cumulant spectrum to build a test statistic that has an asymptotic complex standard normal distribution. We derive the test statistic, study the size and power properties of the test, and demonstrate its implementation with intraday stock market return data. The test has conservative size properties and good power to detect varying variance and unit root in the presence of varying variance.

We suggest a general approach to quantification of different forms of uncertainty in regression tasks performed by artificial neural networks. It is based on the simultaneous training of two neural networks with a joint loss function. One of the networks performs predictions and the other simultaneously quantifies the uncertainty of predictions by estimating the locally averaged loss of the first one. Unlike in many classical uncertainty quantification methods, the targets are not assumed to be sampled from a probability distribution of an a priori given form. We analyze how the hyperparameters affect the learning process and, additionally, show that our method even allows for better predictions compared to standard neural networks without uncertainty counterparts. Finally, we show that particular cases of our approach include maximization of log-likelihood, assuming Gaussian or Laplace noise.

The ungrammatical sentence The key to the cabinets are on the table is known to lead to an illusion of grammaticality. As discussed in the meta-analysis by Jäger et al., 2017, faster reading times are observed at the verb are in the agreement-attraction sentence above compared to the equally ungrammatical sentence The key to the cabinet are on the table. One explanation for this facilitation effect is the feature percolation account: the plural feature on cabinets percolates up to the head noun key, leading to the illusion. An alternative account is in terms of cue-based retrieval account (Lewis & Vasishth, 2005), which assumes that the non-subject noun cabinets is misretrieved due to a partial feature-match when a dependency completion process at the auxiliary initiates a memory access for a subject with plural marking. We present evidence for yet another explanation for the observed facilitation. Because the second sentence has two nouns with identical number, it is possible that these are, in some proportion of trials, more difficult to keep distinct, leading to slower reading times at the verb in the first sentence above; this is the feature overwriting account of Nairne, 1990. We show that the feature overwriting proposal can be implemented as a finite mixture process. We reanalysed ten published data-sets, fitting hierarchical Bayesian mixture models to these data assuming a two-mixture distribution. We show that in nine out of the ten studies, a mixture distribution corresponding to feature overwriting furnishes a superior fit over both the feature percolation and the cue-based retrieval accounts.

The self-concordant-like property of a smooth convex function is a new analytical structure that generalizes the self-concordant notion. While a wide variety of important applications feature the self-concordant-like property, this concept has heretofore remained unexploited in convex optimization. To this end, we develop a variable metric framework of minimizing the sum of a "simple" convex function and a self-concordant-like function. We introduce a new analytic step-size selection procedure and prove that the basic gradient algorithm has improved convergence guarantees as compared to "fast" algorithms that rely on the Lipschitz gradient property. Our numerical tests with real-data sets shows that the practice indeed follows the theory.

Earlier this year, 60,000 technology experts from 170 countries descended on Lisbon, Portugal, to take part in Web Summit, the world's largest tech conference. As part of Web Summit, I attended MoneyConf, an insurtech and fintech conference, where the world's leading insurance companies, banks, tech firms and disruptive startups met. Here, I spoke with industry leaders about how they were changing insurance's long-standing image problem and improving user experience (UX) to better serve customers. Customers are unlikely to do business with any company that doesn't give them a good experience. "Customers want simplicity, more clarity. 'What did I purchase when I signed the contract?'" said Olaf Frank, head of global applications and interim CIO at Munich Re Group, one of the world's leading reinsurers.