IBM says it wants to make intelligent computers that can make decisions like humans. This week, it shipped the NS16e, its largest brain-inspired computer yet, and has big goals ahead. The company plans to create bigger versions of the NS16e -- which was purchased by Lawrence Livermore National Laboratory -- to come closer to matching the scale of a human brain. "Perhaps one day we may see a single rack of neurosynaptic system with as many neurons and synapses as in a human brain," said Jun Sawada, a researcher at IBM, in a blog entry. The brain can be viewed as an extremely power-efficient biological computer.

In July 2013, an up-and-coming artist had an exhibition at the Galerie Oberkampf in Paris. It lasted for a week, was attended by the public, received press coverage, and featured works produced over a number of years, including some created on the spot in the gallery. Altogether, it was a fairly typical art-world event. The only unusual feature was that the artist in question was a computer program known as "The Painting Fool." Even that was not such a novelty.

Structure discovery in graphical models is the determination of the topology of a graph that encodes conditional independence properties of the joint distribution of all variables in the model. For some class of probability distributions, an edge between two variables is present if and only if the corresponding entry in the precision matrix is non-zero. For a finite sample estimate of the precision matrix, entries close to zero may be due to low sample effects, or due to an actual association between variables; these two cases are not readily distinguishable. %Fisher provided a hypothesis test based on a parametric approximation to the distribution of an entry in the precision matrix of a Gaussian distribution, but this may not provide valid upper bounds on $p$-values for non-Gaussian distributions. Many related works on this topic consider potentially restrictive distributional or sparsity assumptions that may not apply to a data sample of interest, and direct estimation of the uncertainty of an estimate of the precision matrix for general distributions remains challenging. Consequently, we make use of results for $U$-statistics and apply them to the covariance matrix. By probabilistically bounding the distortion of the covariance matrix, we can apply Weyl's theorem to bound the distortion of the precision matrix, yielding a conservative, but sound test threshold for a much wider class of distributions than considered in previous works. The resulting test enables one to answer with statistical significance whether an edge is present in the graph, and convergence results are known for a wide range of distributions. The computational complexities is linear in the sample size enabling the application of the test to large data samples for which computation time becomes a limiting factor. We experimentally validate the correctness and scalability of the test on multivariate distributions for which the distributional assumptions of competing tests result in underestimates of the false positive ratio. By contrast, the proposed test remains sound, promising to be a useful tool for hypothesis testing for diverse real-world problems.

This paper is a comment on the survey paper by Biau and Scornet (2016) about random forests. We focus on the problem of quantifying the impact of each ingredient of random forests on their performance. We show that such a quantification is possible for a simple pure forest, leading to conclusions that could apply more generally. Then, we consider "holdout" random forests, which are a good middle point between "toy" pure forests and Breiman's original random forests. We would like to thank G. Biau and E. Scornet for their clear and thought-provoking survey (Biau and Scornet, 2016).

Reciprocal processes are acausal generalizations of Markov processes introduced by Bernstein in 1932. In the literature, a significant amount of attention has been focused on developing dynamical models for reciprocal processes. Recently, probabilistic graphical models for reciprocal processes have been provided. This opens the way to the application of efficient inference algorithms in the machine learning literature to solve the smoothing problem for reciprocal processes. Such algorithms are known to converge if the underlying graph is a tree. This is not the case for a reciprocal process, whose associated graphical model is a single loop network. The contribution of this paper is twofold. First, we introduce belief propagation for Gaussian reciprocal processes. Second, we establish a link between convergence analysis of belief propagation for Gaussian reciprocal processes and stability theory for differentially positive systems.

The back-propagation algorithm is widely used for learning in artificial neural networks. A challenge in machine learning is to create models that generalize to new data samples not seen in the training data. Recently, a common flaw in several machine learning algorithms was discovered: small perturbations added to the input data lead to consistent misclassification of data samples. Samples that easily mislead the model are called adversarial examples. Training a "maxout" network on adversarial examples has shown to decrease this vulnerability, but also increase classification performance. This paper shows that adversarial training has a regularizing effect also in networks with logistic, hyperbolic tangent and rectified linear units. A simple extension to the back-propagation method is proposed, that adds an adversarial gradient to the training. The extension requires an additional forward and backward pass to calculate a modified input sample, or mini batch, used as input for standard back-propagation learning. The first experimental results on MNIST show that the "adversarial back-propagation" method increases the resistance to adversarial examples and boosts the classification performance. The extension reduces the classification error on the permutation invariant MNIST from 1.60% to 0.95% in a logistic network, and from 1.40% to 0.78% in a network with rectified linear units. Results on CIFAR-10 indicate that the method has a regularizing effect similar to dropout in fully connected networks. Based on these promising results, adversarial back-propagation is proposed as a stand-alone regularizing method that should be further investigated.

The task of algorithm selection involves choosing an algorithm from a set of algorithms on a per-instance basis in order to exploit the varying performance of algorithms over a set of instances. The algorithm selection problem is attracting increasing attention from researchers and practitioners in AI. Years of fruitful applications in a number of domains have resulted in a large amount of data, but the community lacks a standard format or repository for this data. This situation makes it difficult to share and compare different approaches effectively, as is done in other, more established fields. It also unnecessarily hinders new researchers who want to work in this area. To address this problem, we introduce a standardized format for representing algorithm selection scenarios and a repository that contains a growing number of data sets from the literature. Our format has been designed to be able to express a wide variety of different scenarios. Demonstrating the breadth and power of our platform, we describe a set of example experiments that build and evaluate algorithm selection models through a common interface. The results display the potential of algorithm selection to achieve significant performance improvements across a broad range of problems and algorithms.

The move from humans working with computers to computers working without humans is almost upon us, and some are already calling it Industry 4.0 – or the fourth industrial revolution. For those not keeping up with your industrial revolutions, the first was considered launched by the use of steam and water power, the second by the use of electricity, and the third by the introduction of computers in the workplace. The name Industry 4.0 was first coined by the German Government, and represents the implementation of artificial intelligence, big data, and the industrial Internet of Things (IIoT) in the factories. It might be the first revolution where humans are not required, according to some. Once computers can talk to each other and automate the assembly line, and AI can understand issues and address them ahead of time, there might be no need for humans.

This is the Delft Dynamics Dronecatcher catching a drone. Robots that hunt other robots are my favorite robots. This is Delft Dynamic's Dronecatcher, a big drone built to hunt smaller drones. Here it is, from the ground. The dronecatcher debuted last year in crude form, and recently Delft Dynamics released new footage of it in action.

The first GTC took place in a set of hotel ballrooms a few blocks away. That's up from 4,000 last year, a growth rate that's tracked pretty steady since the start of the show. The stage is about five feet off the ground. And on the vast screen is an NVIDIA-green moving image that, as it scans looks like a multi-level rendering of the brain's neural network. With some electronics thrown in between. A great many of those here, though, are scientists and analysts of the computational sort -- those who rely on NVIDIA GPUs to help them crunch the rising sea of data that's engulfing us. A lot are associated with universities, close to 200 of them. Virtually every one of the top 100 university comp sci departments are here. There are also hundreds of companies represented--certainly the dozens of major web-services companies that use artificial intelligence. But also industrials, oil and gas, retail. Err, less so this time. But folks don't seem to mind.