Drone users in the UK will be required to do safety awareness tests as part of planned new legislation on their usage. Police will also be given new powers to crack down on illegal use of the unmanned aerial vehicles. The government hopes to harness new drone technology which could see them used on oil rigs, in construction, for organ transport and parcel deliveries. The bill has been welcomed by the pilots' union, which has warned of near misses involving drones and aircraft. Balpa said there had been 81 incidents so far this year - up from 71 in 2016 and 29 in 2015.

In statistical relational learning, knowledge graph completion deals with automatically understanding the structure of large knowledge graphs---labeled directed graphs---and predicting missing relationships---labeled edges. State-of-the-art embedding models propose different trade-offs between modeling expressiveness, and time and space complexity. We reconcile both expressiveness and complexity through the use of complex-valued embeddings and explore the link between such complex-valued embeddings and unitary diagonalization. We corroborate our approach theoretically and show that all real square matrices---thus all possible relation/adjacency matrices---are the real part of some unitarily diagonalizable matrix. This results opens the door to a lot of other applications of square matrices factorization. Our approach based on complex embeddings is arguably simple, as it only involves a Hermitian dot product, the complex counterpart of the standard dot product between real vectors, whereas other methods resort to more and more complicated composition functions to increase their expressiveness. The proposed complex embeddings are scalable to large data sets as it remains linear in both space and time, while consistently outperforming alternative approaches on standard link prediction benchmarks.

UBS Group AG is expanding its workforce in one of the few areas in banking where demand for talent is growing. "We're currently recruiting more people for artificial intelligence," Veronica Lange, head of innovation at Switzerland's biggest bank, said in an interview in Moscow. "These are data scientists, architects, business analysts." AI refers to technology capable of performing tasks that normally require human intelligence. Big global banks like UBS are using it to scour vast databases for insight on customers and markets that could help lenders stay competitive as more and more technology firms delve into financial services.

Could artificial intelligence be the end of the dubious science of connoisseurship? According to a new study, a form of AI called a recurrent neural network may now be able to identify forged paintings. Researchers from New Jersey's Rutgers University and the Atelier for Restoration & Research of Paintings in the Netherlands have published their findings in a paper, titled "Picasso, Matisse, or a Fake? The AI was able to find fake artworks simply by comparing the strokes used to compose the image. Determining the authenticity of a work of art has long been a considerable challenge.

Next to the M56, on the outskirts of Manchester, the future has landed. A cluster of huge distribution centres sits at the heart of Airport City, a new development part-funded by the Beijing Construction Engineering Group (two years ago, it was visited by president Xi Jinping of China). Among the biggest buildings is one of Amazon's self-styled "fulfilment centres". Known within the company as MAN1, it opened in September last year, but everything inside, from the chairs to the wall-mounted screens, looks as if it has just come out of a box. Deeper within the centre, beyond the reception area and meeting rooms, there is something else just as new: a great expanse of space behind a metal cage, where dozens of robots, finished in Amazon orange and each emblazoned with its own number, glide across the floor, gracefully avoiding collisions and sprinting to their next task. Amazon employees call them "drives", but to all intents and purposes these are droids, summoned from the dreams of science fiction and put to work. In some Amazon warehouses, workers – or, in the company's parlance, "associates" – still pace up and down huge aisles, picking out goods and preparing them for shipment; these shifts are said sometimes to involve hikes of 11 miles. The humans in charge of the process known as "picking" now remain in closed workstations, built around a screen that tells them what they need to get next, while the robots bring the shelves – reinvented as four-sided fabric towers, full of pouches that contain everything from DVDs to dolls – to them. There are tasks only a human can do, such as the careful packing of boxes.

Lord Martin Rees, Astronomer Royal and University of Cambridge Emeritus Professor of Cosmology and Astrophysics, believes that machines could surpass humans within a few hundred years, ushering in eons of domination. He also cautions that while we will certainly discover more about the origins of biological life in the coming decades, we should recognize that alien intelligence may be electronic. "Just because there's life elsewhere doesn't mean that there is intelligent life," Lord Rees told The Conversation. "My guess is that if we do detect an alien intelligence, it will be nothing like us. It will be some sort of electronic entity."

On a recent trip to Grenoble, I met with the team behind a "digital nose" that was designed to provide a digital means of detecting smells (you can read about it here). Such digitally augmented sensing is clearly something the area specializes in, as a team of researchers from the GIPSA-Lab in Grenoble has also developed a virtual tongue. The work, which was documented in a recently published paper, uses an ultrasound probe positioned under the jaw with a machine learning algorithm, then takes this data and converts them into virtual replicas in an avatar. The avatar is capable of replicating the movements in the face, the lips, the tongue, and teeth. The researchers believe this visual biofeedback system could provide valuable information for things such as speech therapy.

A new drone designed by Russian researchers is the hulk of the quadcopter world - and can carry a 400-pound (181-kg) payload and fly for up to eight hours. The multi-rotor, autonomous drone, called SKYF, was designed with logistics and agribusinesses companies in mind to create a air freight platform to help business carry out tasks. The vertical take-off and landing drone has applications in areas such as the aerial application of pesticides and fertilizers, seed planting for forest restoration and emergency situations for food and medicine delivery. The drone, designed by Russian company ARDN technology, has a maximum flight speed of 70 kilometers per hour (43.5 miles per hour) at a maximum height of 3,000 meters (9,843 feet) and has a positional accuracy of 30 centimeters (11.8 inches) The drone, designed by Russian company ARDN technology, has a maximum flight speed of 70 kilometers per hour (43.5 miles per hour) and is 5.2 meters (17 feet) by 2.2 meters (7.2 feet). It can fly at a maximum height of 3,000 meters (9,843 feet) and has a positional accuracy of 30 centimeters (11.8 inches).

Kernel methods are powerful tools to capture nonlinear patterns behind data. They implicitly learn high (even infinite) dimensional nonlinear features in the Reproducing Kernel Hilbert Space (RKHS) while making the computation tractable by leveraging the kernel trick. Classic kernel methods learn a single layer of nonlinear features, whose representational power may be limited. Motivated by recent success of deep neural networks (DNNs) that learn multi-layer hierarchical representations, we propose a Stacked Kernel Network (SKN) that learns a hierarchy of RKHS-based nonlinear features. SKN interleaves several layers of nonlinear transformations (from a linear space to a RKHS) and linear transformations (from a RKHS to a linear space). Similar to DNNs, a SKN is composed of multiple layers of hidden units, but each parameterized by a RKHS function rather than a finite-dimensional vector. We propose three ways to represent the RKHS functions in SKN: (1)nonparametric representation, (2)parametric representation and (3)random Fourier feature representation. Furthermore, we expand SKN into CNN architecture called Stacked Kernel Convolutional Network (SKCN). SKCN learning a hierarchy of RKHS-based nonlinear features by convolutional operation with each filter also parameterized by a RKHS function rather than a finite-dimensional matrix in CNN, which is suitable for image inputs. Experiments on various datasets demonstrate the effectiveness of SKN and SKCN, which outperform the competitive methods.

Predicting subgroups of patients with different prognosis is a key challenge for personalized medicine, see for instance Alizadeh et al. [2000] and Rosenwald et al. [2002] where subgroups of patients with different survival rates are identified based on gene expression data. A substantial number of techniques can be found in the literature to predict the subgroup of a given patient in a classification setting, namely when subgroups are known in advance [Golub et al., 1999, Hastie et al., 2001, Tibshirani et al., 2002]. We consider in the present paper the much more difficult case where subgroups are unknown. In this situation, a first widespread approach consists in first using unsupervised learning techniques applied on the covariates - for instance on the gene expression data [Bhattacharjee et al., 2001, Beer et al., 2002, Sørlie et al., 2001] - to define subsets of patients and then estimating the risks in each of them. The problem of such techniques is that there is no guarantee that the identified subgroups will have different risks.