Superior verbal and written communication skills. PREFERRED QUALIFICATIONS Ability to convey rigorous mathematical concepts and considerations to non-experts. Ability to distill problem definitions, models, and constraints from informal business requirements; and to deal with ambiguity and competing objectives.

Everyone seems to have an idea for making them useful--Amazon wants to use them for deliveries, Facebook sees them beaming the Internet to remote areas--but so far it all seems so very outlandish. A California startup called Zipline has a practical plan to use the devices for good. Later this year the company, working with UPS and vaccine distributer Gavi, plans to deploy a fleet of drones in Rwanda, where the machines will deliver medical supplies. The goal is to see 15 autonomous aircraft flying out of a centralized hub make 150 deliveries each day to 21 medical stations throughout the western half of the country. The Rwandan government, which has embraced drone technology and recently approved remarkably progressive guidelines for its use, is working with Zipline to measure the success of the venture.

On a bitter, soul-shivering, damp, biting gray February day in Cleveland--that is to say, on a February day in Cleveland--a handless man is handling a nonexistent ball. Igor Spetic lost his right hand when his forearm was pulped in an industrial accident six years ago and had to be amputated. In an operation four years ago, a team of surgeons implanted a set of small translucent "interfaces" into the neural circuits of his upper arm. This afternoon, in a basement lab at a Veterans Administration hospital, the wires are hooked up directly to a prosthetic hand--plastic, flesh-colored, five-fingered, and articulated--that is affixed to what remains of his arm. The hand has more than a dozen pressure sensors within it, and their signals can be transformed by a computer into electric waves like those natural to the nervous system. Since, from the brain's point of view, his hand is still there, it needs only to be recalled to life. With the "stimulation" turned on--the electronic feed coursing from the sensors--Spetic feels nineteen distinct sensations in his artificial hand. Above all, he can feel pressure as he would with a living hand. "We don't appreciate how much of our behavior is governed by our intense sensitivity to pressure," Dustin Tyler, the fresh-faced principal investigator on the Cleveland project, says, observing Spetic closely. "We think of hot and cold, or of textures, silk and cotton. But some of the most important sensing we do with our fingers is to register incredibly minute differences in pressure, of the kinds that are necessary to perform tasks, which we grasp in a microsecond from the feel of the outer shell of the thing. We know instantly, just by touching, whether to gently squeeze the toothpaste or crush the can." With the new prosthesis, Spetic can sense the surface of a cherry in a way that allows him to stem it effortlessly and precisely, guided by what he feels, rather than by what he sees. Prosthetic hands like Spetic's tend to be super-strong, capable of forty pounds of pressure, so the risk of crushing an egg is real. The stimulation sensors make delicate tasks easy. Spetic comes into the lab every other week; the rest of the time he is busy pursuing a degree in engineering, which he has taken up while on disability.

It's no secret that Google has broad ambitions in healthcare. But a document obtained by New Scientist reveals that the tech giant's collaboration with the UK's National Health Service goes far beyond what has been publicly announced. The document – a data-sharing agreement between Google-owned artificial intelligence company DeepMind and the Royal Free NHS Trust – gives the clearest picture yet of what the company is doing and what sensitive data it now has access to. The agreement gives DeepMind access to a wide range of healthcare data on the 1.6 million patients who pass through three London hospitals run by the Royal Free NHS Trust – Barnet, Chase Farm and the Royal Free – each year. This will include information about people who are HIV-positive, for instance, as well as details of drug overdoses and abortions.

Efficient and accurate low-rank approximations of multiple data sources are essential in the era of big data. The scaling of kernel-based learning algorithms to large datasets is limited by the O(n^2) computation and storage complexity of the full kernel matrix, which is required by most of the recent kernel learning algorithms. We present the Mklaren algorithm to approximate multiple kernel matrices learn a regression model, which is entirely based on geometrical concepts. The algorithm does not require access to full kernel matrices yet it accounts for the correlations between all kernels. It uses Incomplete Cholesky decomposition, where pivot selection is based on least-angle regression in the combined, low-dimensional feature space. The algorithm has linear complexity in the number of data points and kernels. When explicit feature space induced by the kernel can be constructed, a mapping from the dual to the primal Ridge regression weights is used for model interpretation. The Mklaren algorithm was tested on eight standard regression datasets. It outperforms contemporary kernel matrix approximation approaches when learning with multiple kernels. It identifies relevant kernels, achieving highest explained variance than other multiple kernel learning methods for the same number of iterations. Test accuracy, equivalent to the one using full kernel matrices, was achieved with at significantly lower approximation ranks. A difference in run times of two orders of magnitude was observed when either the number of samples or kernels exceeds 3000.

The refugee crisis is perhaps the single most challenging problem for Europe today. Hundreds of thousands of people have already traveled across dangerous sea passages from Turkish shores to Greek islands, resulting in thousands of dead and missing, despite the best rescue efforts from both sides. One of the main reasons is the total lack of any early warning-alerting system, which could provide some preparation time for the prompt and effective deployment of resources at the hot zones. This work is such an attempt for a systemic analysis of the refugee influx in Greece, aiming at (a) the statistical and signal-level characterization of the smuggling networks and (b) the formulation and preliminary assessment of such models for predictive purposes, i.e., as the basis of such an early warning-alerting protocol. To our knowledge, this is the first-ever attempt to design such a system, since this refugee crisis itself and its geographical properties are unique (intense event handling, little or no warning). The analysis employs a wide range of statistical, signal-based and matrix factorization (decomposition) techniques, including linear & linear-cosine regression, spectral analysis, ARMA, SVD, Probabilistic PCA, ICA, K-SVD for Dictionary Learning, as well as fractal dimension analysis. It is established that the behavioral patterns of the smuggling networks closely match (as expected) the regular burst and pause periods of store-and-forward networks in digital communications. There are also major periodic trends in the range of 6.2-6.5 days and strong correlations in lags of four or more days, with distinct preference in the Sunday-Monday 48-hour time frame. These results show that such models can be used successfully for short-term forecasting of the influx intensity, producing an invaluable operational asset for planners, decision-makers and first-responders.

Data representation in a stacked denoising autoencoder is investigated. Decoding is a simple technique for translating a stacked denoising autoencoder into a composition of denoising autoencoders in the ground space. In the infinitesimal limit, a composition of denoising autoencoders is reduced to a continuous denoising autoencoder, which is rich in analytic properties and geometric interpretation. For example, the continuous denoising autoencoder solves the backward heat equation and transports each data point so as to decrease entropy of the data distribution. Together with ridgelet analysis, an integral representation of a stacked denoising autoencoder is derived.

Neural signals are characterized by rich temporal and spatiotemporal dynamics that reflect the organization of cortical networks. Theoretical research has shown how neural networks can operate at different dynamic ranges that correspond to specific types of information processing. Here we present a data analysis framework that uses a linearized model of these dynamic states in order to decompose the measured neural signal into a series of components that capture both rhythmic and non-rhythmic neural activity. The method is based on stochastic differential equations and Gaussian process regression. Through computer simulations and analysis of magnetoencephalographic data, we demonstrate the efficacy of the method in identifying meaningful modulations of oscillatory signals corrupted by structured temporal and spatiotemporal noise. These results suggest that the method is particularly suitable for the analysis and interpretation of complex temporal and spatiotemporal neural signals.

This article addresses the problem of derivative-free (single- or multi-objective) optimization subject to multiple inequality constraints. Both the objective and constraint functions are assumed to be smooth, non-linear and expensive to evaluate. As a consequence, the number of evaluations that can be used to carry out the optimization is very limited, as in complex industrial design optimization problems. The method we propose to overcome this difficulty has its roots in both the Bayesian and the multi-objective optimization literatures. More specifically, an extended domination rule is used to handle objectives and constraints in a unified way, and a corresponding expected hyper-volume improvement sampling criterion is proposed. This new criterion is naturally adapted to the search of a feasible point when none is available, and reduces to existing Bayesian sampling criteria---the classical Expected Improvement (EI) criterion and some of its constrained/multi-objective extensions---as soon as at least one feasible point is available. The calculation and optimization of the criterion are performed using Sequential Monte Carlo techniques. In particular, an algorithm similar to the subset simulation method, which is well known in the field of structural reliability, is used to estimate the criterion. The method, which we call BMOO (for Bayesian Multi-Objective Optimization), is compared to state-of-the-art algorithms for single- and multi-objective constrained optimization.

The first commercial spam message was sent in 1994--at least that's the general consensus. Lawrence Canter and Margaret Siegel had a program written that would post a copy of an advertisement for their law firm's green card lottery paperwork service to every Usenet news group -- about 6,000 of them. Because of the way the messages were posted, Usenet clients couldn't filter out duplicate copies, and users saw a copy of the same message in every group. At the time, commercial use of internet resources was rare (it had only recently become legal) and access to Usenet was expensive. Users considered these commercial-seeming messages to be crass--not only did they take up their time, but they also cost them money.