Christophe Deschamps was watching a basketball game with his wife and three children when he received an alert on his smartphone. The home security system told him something was wrong, so he quickly accessed the video feed on his phone. "I could see smoke," he says. Their home, in the Wallonia region of southern Belgium, was on fire. The family's thoughts immediately turned to their two Bernese Mountain dogs - Lisbonne and Hawaii - locked in the garage.

A team of researchers at the University of Nottingham, UK have developed a system that uses AI to predict heart attacks better than doctors. This machine-learning algorithm can now predict the likelihood of heart attacks and strokes in people just as any doctor, only more accurately. As reported in Science, this new method employs artificial intelligence (AI) is scientifically proven to be capable of performing better at studying cardiovascular disease than the standard medical guideline. After a number of studies, The American College of Cardiology/American Heart Association (ACC/AHA) has developed a series of guideline based on eight primary factors such as age, blood pressure and cholesterol level with four different machine-learning languages. This system is believed to predict heart risk by 72.8 percent. Stephen Weng and his team at the University of Nottingham fed the data of 378,256 patients from the UK, in which about 295,000 were first used for generating their initial predictive model.

Information systems experience an ever-growing volume of unstructured data, particularly in the form of textual materials. This represents a rich source of information from which one can create value for people, organizations and businesses. For instance, recommender systems can benefit from automatically understanding preferences based on user reviews or social media. However, it is difficult for computer programs to correctly infer meaning from narrative content. One major challenge is negations that invert the interpretation of words and sentences. As a remedy, this paper proposes a novel learning strategy to detect negations: we apply reinforcement learning to find a policy that replicates the human perception of negations based on an exogenous response, such as a user rating for reviews. Our method yields several benefits, as it eliminates the former need for expensive and subjective manual labeling in an intermediate stage. Moreover, the inferred policy can be used to derive statistical inferences and implications regarding how humans process and act on negations.

This paper describes a new neuroimaging analysis toolbox that allows for the modeling of nonlinear effects at the voxel level, overcoming limitations of methods based on linear models like the GLM. We illustrate its features using a relevant example in which distinct nonlinear trajectories of Alzheimer's disease related brain atrophy patterns were found across the full biological spectrum of the disease. The open-source toolbox presented in this paper is available at https://github.com/imatge-upc/VNeAT.

In an era where accumulating data is easy and storing it inexpensive, feature selection plays a central role in helping to reduce the high-dimensionality of huge amounts of otherwise meaningless data. In this paper, we propose a graph-based method for feature selection that ranks features by identifying the most important ones into arbitrary set of cues. Mapping the problem on an affinity graph-where features are the nodes-the solution is given by assessing the importance of nodes through some indicators of centrality, in particular, the Eigen-vector Centrality (EC). The gist of EC is to estimate the importance of a feature as a function of the importance of its neighbors. Ranking central nodes individuates candidate features, which turn out to be effective from a classification point of view, as proved by a thoroughly experimental section. Our approach has been tested on 7 diverse datasets from recent literature (e.g., biological data and object recognition, among others), and compared against filter, embedded and wrappers methods. The results are remarkable in terms of accuracy, stability and low execution time.

The challenge of taking many variables into account in optimization problems may be overcome under the hypothesis of low effective dimensionality. Then, the search of solutions can be reduced to the random embedding of a low dimensional space into the original one, resulting in a more manageable optimization problem. Specifically, in the case of time consuming black-box functions and when the budget of evaluations is severely limited, global optimization with random embeddings appears as a sound alternative to random search. Yet, in the case of box constraints on the native variables, defining suitable bounds on a low dimensional domain appears to be complex. Indeed, a small search domain does not guarantee to find a solution even under restrictive hypotheses about the function, while a larger one may slow down convergence dramatically. Here we tackle the issue of low-dimensional domain selection based on a detailed study of the properties of the random embedding, giving insight on the aforementioned difficulties. In particular, we describe a minimal low-dimensional set in correspondence with the embedded search space. We additionally show that an alternative equivalent embedding procedure yields simultaneously a simpler definition of the low-dimensional minimal set and better properties in practice. Finally, the performance and robustness gains of the proposed enhancements for Bayesian optimization are illustrated on three examples.

Last week, I spoke at the Swiss Mobile Association. The event was held at one of the oldest cross-functional research institutes Gottlieb Duttweiler Institute just outside Zurich. Prior to being involved in IoT and AI, I worked for many years in Telecoms. I believe that from an innovation standpoint – we are living in a post-mobile world. Today, just as the Web itself, Mobile is a mature industry.

To keep contraband out of prisons, the United Kingdom wants drones out of the sky. The nation announced a new "squad" on Monday that will trace captured drones back to their owners – something like a group of detectives. Working with "national law enforcement agencies and HM Prison and Probation Service," the new squad will try to match captured drones with their owners so they can figure out who is trying to smuggle drugs, cellphones and other items over prison walls. Smugglers used drones to get contraband over prison walls 33 times in 2015 compared with just two times in 2014 and none in 2013, according to the Press Association. UK police have thrown drone smugglers in jail for several years after recent convictions.

A little over two years ago, Russia's largest search provider, Yandex, filed a complaint against Google for what it believed were anti-competitive practices. Now, Russia's Federal Antimonopoly Service has given credence to Yandex's claims and "issued a prescription to Google in order to require the company to remove anti-competitive restrictions from its agreements with manufacturers," according to a press release. That means Google won't be able to pre-install apps on phones, control the default search engine or place its own apps on device home screens in the region. Furthermore: "Google will be committed to securing the rights of the third parties to include their search engines in the choice window." In a "few months," Google will have a home-screen search widget that will offer up any manner of search providers (yep, including Yandex) assuming they sign a commercial agreement for their inclusion in the query box. For its part, the FAS says that this is a good move to ensure everyone is on the same footing in terms of competition for app placement and web searches.

Between 15 and 20 million people die every year from heart attacks and related illnesses worldwide, but now, artificial intelligence could help reduce that number with better predictive abilities. Doctors are not clairvoyant, but it looks like technology is getting awfully close. Thanks to a team of researchers at the University of Nottingham in the United Kingdom, we could be closer than ever before to predicting the future when it comes to patients' health risks. The scientists have managed to develop an algorithm that outperforms medical doctors when it comes to predicting heart attacks. And this, experts say, could save thousands or even millions of lives every year.