Traditional dictionary learning methods are based on quadratic convex loss function and thus are sensitive to outliers. In this paper, we propose a generic framework for robust dictionary learning based on concave losses. We provide results on composition of concave functions, notably regarding super-gradient computations, that are key for developing generic dictionary learning algorithms applicable to smooth and non-smooth losses. In order to improve identification of outliers, we introduce an initialization heuristic based on undercomplete dictionary learning. Experimental results using synthetic and real data demonstrate that our method is able to better detect outliers, is capable of generating better dictionaries, outperforming state-of-the-art methods such as K-SVD and LC-KSVD.

Bill Schneider, vice president of IT at Pioneer Energy Services, said oil and gas has historically been "a laggard in digital." And tech investments slowed in 2014, when the price of oil started its precipitous, two-year decline. "So we've got a lot of ground to make up," Schneider said. The San Antonio-based company provides drilling and well services for oil and gas companies in the U.S. and Colombia, and it has sensors affixed to wells and field equipment "which generate a tremendous amount of data," Schneider said. But just a fraction of the data coursing through the internet of things (IoT) and collected by the company is analyzed, he said -- and that presents a huge opportunity.

Participatory democracy advances in virtually all governments and especially in South America which exhibits a mixed culture and social predisposition. This article presents the "Social Participation Ontology" (OPS from the Brazilian name \emph{Ontologia de Participa\c{c}\~ao Social}) implemented in compliance with the Web Ontology Language standard (OWL) for fostering social participation, specially in virtual platforms. The entities and links of OPS were defined based on an extensive collaboration of specialists. It is shown that OPS is instrumental for information retrieval from the contents of the portal, both in terms of the actors (at various levels) as well as mechanisms and activities. Significantly, OPS is linked to other OWL ontologies as an upper ontology and via FOAF and BFO as higher upper ontologies, which yields sound organization and access of knowledge and data. In order to illustrate the usefulness of OPS, we present results on ontological expansion and integration with other ontologies and data. Ongoing work involves further adoption of OPS by the official Brazilian federal portal for social participation and NGO s, and further linkage to other ontologies for social participation.

Late in 2015, Gilberto Titericz, an electrical engineer at Brazil's state oil company Petrobras, told his boss he planned to resign, after seven years maintaining sensors and other hardware in oil plants. By devoting hundreds of hours of leisure time to the obscure world of competitive data analysis, Titericz had recently become the world's top-ranked data scientist, by one reckoning. "Only when I wanted to quit did they realize they had the number-one data scientist," he says. Petrobras held on to its champ for a time by moving Titericz into a position that used his data skills. But since topping the rankings that October he'd received a stream of emails from recruiters around the globe, including representatives of Tesla and Google.

Fast-food workers, cashiers, cooks, delivery people and their supporters held a rally outside New York City Hall on May 24, 2017.Erik Mcgregor/Pacific Press/Zuma From the window of his university office in Louvain-la-Neuve, Belgium, philosophy professor Philippe Van Parijs--considered by many to be Europe's most prominent advocate for the idea that the state should provide a regular income to every citizen--can see the mailbox where he sent off invitations to the first "basic income" conference more than 30 years ago. "I'm quite amazed by the seed we threw on the ground now," he says. After decades of obscurity, the idea is suddenly in fashion. Politicians around the world are interested and a handful of governments, such as Finland and the Canadian province of Ontario, are planning or considering basic-income pilot projects. But the idea of basic income has been around for more than 200 years, rising on waves of political and economic turmoil only to disappear in calmer times.

New forms of systems of intelligence are emerging through embedded artificial intelligence, machine learning, and analytics. These data-driven systems of intelligence are enabling digital disruption and new business models. Many companies don't know what steps to take to become digital, where to begin their journey to digital, or how to be sure they won't waste money on innovation they can't implement throughout their company to drive better business results. There is a massive opportunity to help companies take and complete this digital journey, not just to innovate, but to become scaled digital businesses. In this Data Science Central webinar join David Judge, Vice President, Chief Evangelist Leonardo at SAP, Bill Vorhies, Data Scientist, Editorial Director of Data Science Central, and Guilherme Rabello, Commercial and Market Intelligence Manager of InovaInCor, the Innovation department of the Heart Institute (InCor) in São Paulo as they discuss how new technologies are driving digital disruption and the need for innovation.

Efficient and high-fidelity prior sampling and inversion for complex geological media is still a largely unsolved challenge. Here, we use a deep neural network of the variational autoencoder type to construct a parametric low-dimensional base model parameterization of complex binary geological media. For inversion purposes, it has the attractive feature that random draws from an uncorrelated standard normal distribution yield model realizations with spatial characteristics that are in agreement with the training set. In comparison with the most commonly used parametric representations in probabilistic inversion, we find that our dimensionality reduction (DR) approach outperforms principle component analysis (PCA), optimization-PCA (OPCA) and discrete cosine transform (DCT) DR techniques for unconditional geostatistical simulation of a channelized prior model. For the considered examples, important compression ratios (200 - 500) are achieved. Given that the construction of our parameterization requires a training set of several tens of thousands of prior model realizations, our DR approach is more suited for probabilistic (or deterministic) inversion than for unconditional (or point-conditioned) geostatistical simulation. Probabilistic inversions of 2D steady-state and 3D transient hydraulic tomography data are used to demonstrate the DR-based inversion. For the 2D case study, the performance is superior compared to current state-of-the-art multiple-point statistics inversion by sequential geostatistical resampling (SGR). Inversion results for the 3D application are also encouraging.

Data and knowledge representation are fundamental concepts in machine learning. The quality of the representation impacts the performance of the learning model directly. Feature learning transforms or enhances raw data to structures that are effectively exploited by those models. In recent years, several works have been using complex networks for data representation and analysis. However, no feature learning method has been proposed for such category of techniques. Here, we present an unsupervised feature learning mechanism that works on datasets with binary features. First, the dataset is mapped into a feature--sample network. Then, a multi-objective optimization process selects a set of new vertices to produce an enhanced version of the network. The new features depend on a nonlinear function of a combination of preexisting features. Effectively, the process projects the input data into a higher-dimensional space. To solve the optimization problem, we design two metaheuristics based on the lexicographic genetic algorithm and the improved strength Pareto evolutionary algorithm (SPEA2). We show that the enhanced network contains more information and can be exploited to improve the performance of machine learning methods. The advantages and disadvantages of each optimization strategy are discussed.

The Global Service Robotics Market is a strategy report from Berg Insight analysing the latest developments on this market covering floor cleaning robots, robot lawn mowers, milking robots, humanoid robots, telepresence robots, powered human exoskeletons, surgical robots, AGVs, AMRs and UAVs. This strategic research report from Berg Insight provides you with 240 pages of unique business intelligence including 5-year industry forecasts and expert commentary on which to base your business decisions. The future is here: Service robotics will change our lives Robots are now being increasingly adopted for service applications, both by consumers and professionals. The service robot market comprises many different types of robots, most of which can be used for applications in multiple industries. At a consumer level service robots are commonly used for tedious and repetitive tasks such as domestic chores, or for leisure and entertainment purposes.

You may think that humans as a species are the most prolific colonizers on the planet but maybe you've never heard of the "Argentine ants of southern Europe". This massive colony is made up of ants from Argentina that were introduced to Europe 80 years ago, and needless to say they've wasted little time in dominating the landscape. While Europe is being invaded by immigrant ants, another much smaller infestation that's taking place is that of mythical unicorns. For those of you not in the know, a unicorn is a startup worth at least $1 billion, and there are 215 in total around the world. There are actually 16 unicorns grazing across Europe right now, and we got zee Germans out of the way after last week's top 8 unicorns by funding.