We summarize the potential impact that the European Union’s new General Data Protection Regulation will have on the routine use of machine learning algorithms. Slated to take effect as law across the EU in 2018, it will restrict automated individual decision-making (that is, algorithms that make decisions based on user-level predictors) which “significantly affect” users. The law will also effectively create a “right to explanation,” whereby a user can ask for an explanation of an algorithmic decision that was made about them. We argue that while this law will pose large challenges for industry, it highlights opportunities for computer scientists to take the lead in designing algorithms and evaluation frameworks which avoid discrimination and enable explanation.

The field of computational models of argument is emerging as an important aspect of artificial intelligence research. The reason for this is based on the recognition that if we are to develop robust intelligent systems, then it is imperative that they can handle incomplete and inconsistent information in a way that somehow emulates the way humans tackle such a complex task. And one of the key ways that humans do this is to use argumentation either internally, by evaluating arguments and counterarguments‚ or externally, by for instance entering into a discussion or debate where arguments are exchanged. As we report in this review, recent developments in the field are leading to technology for artificial argumentation, in the legal, medical, and e-government domains, and interesting tools for argument mining, for debating technologies, and for argumentation solvers are emerging.

Recent advances in artificial intelligence are encouraging governments and corporations to deploy AI in high-stakes settings including driving cars autonomously, managing the power grid, trading on stock exchanges, and controlling autonomous weapons systems. Such applications require AI methods to be robust to both the known unknowns (those uncertain aspects of the world about which the computer can reason explicitly) and the unknown unknowns (those aspects of the world that are not captured by the system’s models). This article discusses recent progress in AI and then describes eight ideas related to robustness that are being pursued within the AI research community. While these ideas are a start, we need to devote more attention to the challenges of dealing with the known and unknown unknowns. These issues are fascinating, because they touch on the fundamental question of how finite systems can survive and thrive in a complex and dangerous world

Studying general quantum many-body systems is one of the major challenges in modern physics because it requires an amount of computational resources that scales exponentially with the size of the system.Simulating the evolution of a state, or even storing its description, rapidly becomes intractable for exact classical algorithms. Recently, machine learning techniques, in the form of restricted Boltzmann machines, have been proposed as a way to efficiently represent certain quantum states with applications in state tomography and ground state estimation. Here, we introduce a new representation of states based on variational autoencoders. Variational autoencoders are a type of generative model in the form of a neural network. We probe the power of this representation by encoding probability distributions associated with states from different classes. Our simulations show that deep networks give a better representation for states that are hard to sample from, while providing no benefit for random states. This suggests that the probability distributions associated to hard quantum states might have a compositional structure that can be exploited by layered neural networks. Specifically, we consider the learnability of a class of quantum states introduced by Fefferman and Umans. Such states are provably hard to sample for classical computers, but not for quantum ones, under plausible computational complexity assumptions. The good level of compression achieved for hard states suggests these methods can be suitable for characterising states of the size expected in first generation quantum hardware.

The use of noninvasive functional Magnetic Resonance Imaging (fMRI) techniques for determining brain activity requires a set of data processing steps that transforms raw data into validated elements suitable for statistical analysis. Gaussian filter to average local voxel intensities is an important preprocessing step. Spatial smoothing serves several purposes [1]: local averaging reduces uncorrelated random noise in the voxel, hence increasing the Signal-to-Noise Ratio (SNR) leading to improved statistical power to detect true functional brain activation; also, spatial smoothing serves to eliminate unimportant anatomical details across subjects, that are preserved despite affine and nonlinear transformations, which are common spatial normalisation steps in the pipeline; additionally, smoothing can ensure that assumptions typically made to enable multiple comparison correction using Random Field Theory (RFT) for locating brain activation are fulfilled. Notice that the resolution of each brain volume is decreased by applying the smoothing step, meaning that an appropriate tradeoff between the original volume and the degree of smoothing to be applied is sought. This compromise is governed by a single parameter stating the width of the Gaussian filter. In spite of the importance of this parameter, there is no established method to automatically select its most appropriate value for every situation, being often set according to best practices or relying on each scientist's expertise.

We present a new method for forecasting systems of multiple interrelated time series. The method learns the forecast models together with discovering leading indicators from within the system that serve as good predictors improving the forecast accuracy and a cluster structure of the predictive tasks around these. The method is based on the classical linear vector autoregressive model (VAR) and links the discovery of the leading indicators to inferring sparse graphs of Granger causality. We formulate a new constrained optimisation problem to promote the desired sparse structures across the models and the sharing of information amongst the learning tasks in a multi-task manner. We propose an algorithm for solving the problem and document on a battery of synthetic and real-data experiments the advantages of our new method over baseline VAR models as well as the state-of-the-art sparse VAR learning methods.

When solving consensus optimization problems over a graph, there is often an explicit characterization of the convergence rate of Gradient Descent (GD) using the spectrum of the graph Laplacian. The same type of problems under the Alternating Direction Method of Multipliers (ADMM) are, however, poorly understood. For instance, simple but important non-strongly-convex consensus problems have not yet being analyzed, especially concerning the dependency of the convergence rate on the graph topology. Recently, for a non-strongly-convex consensus problem, a connection between distributed ADMM and lifted Markov chains was proposed, followed by a conjecture that ADMM is faster than GD by a square root factor in its convergence time, in close analogy to the mixing speedup achieved by lifting several Markov chains. Nevertheless, a proof of such a claim is is still lacking. Here we provide a full characterization of the convergence of distributed over-relaxed ADMM for the same type of consensus problem in terms of the topology of the underlying graph. Our results provide explicit formulas for optimal parameter selection in terms of the second largest eigenvalue of the transition matrix of the graph's random walk. Another consequence of our results is a proof of the aforementioned conjecture, which interestingly, we show it is valid for any graph, even the ones whose random walks cannot be accelerated via Markov chain lifting.

We investigate crowdsourcing algorithms for finding the top-quality item within a large collection of objects with unknown intrinsic quality values. This is an important problem with many relevant applications, for example in networked recommendation systems. The core of the algorithms is that objects are distributed to crowd workers, who return a noisy and biased evaluation. All received evaluations are then combined, to identify the top-quality object. We first present a simple probabilistic model for the system under investigation. Then, we devise and study a class of efficient adaptive algorithms to assign in an effective way objects to workers. We compare the performance of several algorithms, which correspond to different choices of the design parameters/metrics. In the simulations we show that some of the algorithms achieve near optimal performance for a suitable setting of the system parameters.

A survey by the McKinsey Global Institute of 3,000 C-level executives across 10 countries and 14 sectors identified five fundamental strategies for how to get the most out of AI's potential, write Jacques Bughin and Eric Hazan in MIT Sloan Management Review. Fueled by the buzz around powerful applications of artificial intelligence (AI), many business leaders are contemplating whether to introduce AI into their organizations. While practitioners and academics have outlined some of the strategic challenges of implementing AI, many executives are still seeking good models for how to generate competitive advantage from its application. To find out more about what contributes to successful AI adoption, we helped lead a survey by the McKinsey Global Institute of 3,000 C-level executives across 10 countries and 14 sectors. From that research, we identified five fundamental strategies for how to get the most out of AI's potential.

Using AI and deep learning to create a quick report on the statistics and quotes used should be relatively simple, scanning through complex and variable data sources to discover patterns that show whether information is correct or not. These tools could even check that images accompanying the article show the correct picture and context. A recent article on Breitbart, for instance, could see them end up in court after they used an image of Lukas Podolski, a German soccer player who has appeared for his country 130 times. The image of Podolski and another man appeared under the headline'Spanish police crack gang moving migrants on jet skis', but a 10 second Google image search would have shown that this was actually an image of Podolski on a Jet Ski trip during the Rio 2016 World Cup. A simple AI system would have picked this up almost instantly through image recognition and allowed them to avoid the embarrassment and potential law suit that Podolski is reportedly considering against them.