We study sequential change-point detection using sketches (linear projections) of high-dimensional signal vectors, by presenting the sketching procedures that are derived based on the generalized likelihood ratio statistic. We consider both fixed and time-varying projections, and derive theoretical approximations to two fundamental performance metrics: the average run length (ARL) and the expected detection delay (EDD); these approximations are shown to be highly accurate by numerical simulations. We also characterize the performance of the procedure when the projection is a Gaussian random projection or a sparse 0-1 matrix (in particular, an expander graph). Finally, we demonstrate the good performance of the sketching performance using simulation and real-data examples on solar flare detection and failure detection in power networks.

This paper proposes a new interpretation of sparse penalties such as the elastic-net and the group-lasso. Beyond providing a new viewpoint on these penalization schemes, our approach results in a unified optimization strategy. Our experiments demonstrate that this strategy, implemented on the elastic-net, is computationally extremely efficient for small to medium size problems. Our accompanying software solves problems very accurately, at machine precision, in the time required to get a rough estimate with competing state-of-the-art algorithms. We illustrate on real and artificial datasets that this accuracy is required to for the correctness of the support of the solution, which is an important element for the interpretability of sparsity-inducing penalties.

In this paper we propose a scalable version of a state-of-the-art deterministic time-invariant feature extraction approach based on consecutive changes of basis and nonlinearities, namely, the scattering network. The first focus of the paper is to extend the scattering network to allow the use of higher order nonlinearities as well as extracting nonlinear and Fourier based statistics leading to the required invariants of any inherently structured input. In order to reach fast convolutions and to leverage the intrinsic structure of wavelets, we derive our complete model in the Fourier domain. In addition of providing fast computations, we are now able to exploit sparse matrices due to extremely high sparsity well localized in the Fourier domain. As a result, we are able to reach a true linear time complexity with inputs in the Fourier domain allowing fast and energy efficient solutions to machine learning tasks. Validation of the features and computational results will be presented through the use of these invariant coefficients to perform classification on audio recordings of bird songs captured in multiple different soundscapes. In the end, the applicability of the presented solutions to deep artificial neural networks is discussed.

We consider the task of one-shot learning of visual categories, or more generally, learning to classify images with few examples of particular classes. The currently dominant image classification paradigm of supervised deep learning performs well only when data is abundant. In this paper we explore a Bayesian procedure for updating a pretrained convnet to classify a novel image category for which data is limited. We demonstrate that the approach is competitive with state-of-the-art methods whilst also being consistent with'normal' methods for training deep networks on large data. Several approaches to one-shot learning have been noted as failing to beat a simple nearest-neighbour classifier [8]. Recent approaches of the problem have used relatively complicated architectures such as memory augmented neural networks [9, 10] or siamese networks [5]; or have been specialised for the task of one-shot learning [10]. Fei-Fei et al. [2] demonstrated one-shot learning as a Bayesian update to an image classification model with a prior based on categories learned with lots of data. Our work is an modern update of this work, applying this technique to deep convolutional networks.

We demonstrate the potential of Deep Learning methods for measurements of cosmological parameters from density fields, focusing on the extraction of non-Gaussian information. We consider weak lensing mass maps as our dataset. We aim for our method to be able to distinguish between five models, which were chosen to lie along the $\sigma_8$ - $\Omega_m$ degeneracy, and have nearly the same two-point statistics. We design and implement a Deep Convolutional Neural Network (DCNN) which learns the relation between five cosmological models and the mass maps they generate. We develop a new training strategy which ensures the good performance of the network for high levels of noise. We compare the performance of this approach to commonly used non-Gaussian statistics, namely the skewness and kurtosis of the convergence maps. We find that our implementation of DCNN outperforms the skewness and kurtosis statistics, especially for high noise levels. The network maintains the mean discrimination efficiency greater than $85\%$ even for noise levels corresponding to ground based lensing observations, while the other statistics perform worse in this setting, achieving efficiency less than $70\%$. This demonstrates the ability of CNN-based methods to efficiently break the $\sigma_8$ - $\Omega_m$ degeneracy with weak lensing mass maps alone. We discuss the potential of this method to be applied to the analysis of real weak lensing data and other datasets.

Competitive methods for multi-label classification typically invest in learning labels together. To do so in a beneficial way, analysis of label dependence is often seen as a fundamental step, separate and prior to constructing a classifier. Some methods invest up to hundreds of times more computational effort in building dependency models, than training the final classifier itself. We extend some recent discussion in the literature and provide a deeper analysis, namely, developing the view that label dependence is often introduced by an inadequate base classifier, rather than being inherent to the data or underlying concept; showing how even an exhaustive analysis of label dependence may not lead to an optimal classification structure. Viewing labels as additional features (a transformation of the input), we create neural-network inspired novel methods that remove the emphasis of a prior dependency structure. Our methods have an important advantage particular to multi-label data: they leverage labels to create effective units in middle layers, rather than learning these units from scratch in an unsupervised fashion with gradient-based methods. Results are promising. The methods we propose perform competitively, and also have very important qualities of scalability.

Essential, the consumer electronics startup from Google Android creator Andy Rubin, is undergoing some executive turnover. This seems to be even more confirmation that the Essential Phone release date is still murky. Brian Wallace, vice president of marketing at Essential, left the company last week, according to Business Insider. Wallace has taken on a role at i.am. Andy Fouché, an advisor who also worked as Essential's head of communications, left the company last month, according to Business Insider.

We propose a fast inference method for Bayesian nonlinear support vector machines that leverages stochastic variational inference and inducing points. Our experiments show that the proposed method is faster than competing Bayesian approaches and scales easily to millions of data points. It provides additional features over frequentist competitors such as accurate predictive uncertainty estimates and automatic hyperparameter search.

Mark Minevich is the principle founder of Going Global Ventures and venture partner of GVA Capital in Silicon Valley. While the public is becoming aware of the possibilities of artificial intelligence, I've been evangelizing, researching and working with it for years. As a fellow of the U.S. Council on Competitiveness, a senior advisor to the United Nations Office for Project Services, and a member of both the World Artificial Intelligence Organization and B20, I've been researching AI (and sharing what I learn) for years. Although AI buzz typically surrounds statements from famous minds like Stephen Hawking or Elon Musk – both of whom rightfully worry about granting sentience to machines that control our way of life -- there are plenty of worthy conversations being had on the subject. AI also represents cybersecurity's next evolution, both as a threat and a solution.

Helping Wimbledon in their pursuit of greatness For 28 years IBM has been the official supplier of Information Technology and consultant to the All England Club and The Championships, Wimbledon. By analysing millions of data points and thousands of pages of unstructured text you can understand, reason and learn what it takes to pursue greatness within a sporting arena. For a business wanting to understand how to deliver a great customer experience or differentiate their services they must look beyond pure historic structured data to the vast volumes of unstructured data created every day. IBM's cognitive and cloud solutions businesses can understand and respond to their customer needs in ways never possible before. This year IBM has used these technologies to determine what makes a great Wimbledon Champion (watch video above).