Convolution Neural Network (CNN) has gained tremendous success in computer vision tasks with its outstanding ability to capture the local latent features. Recently, there has been an increasing interest in extending CNNs to the general spatial domain. Although various types of graph and geometric convolution methods have been proposed, their connections to traditional 2D-convolution are not well-understood. In this paper, we show that depthwise separable convolution is the key to close the gap, based on which we derive a novel Depthwise Separable Graph Convolution that subsumes existing graph convolution methods as special cases of our formulation. Experiments show that the proposed approach consistently outperforms other graph and geometric convolution baselines on benchmark datasets in multiple domains.

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.

Over-subscription, that is, being assigned too many things to do, is commonly encountered in temporal scheduling problems. As human beings, we often want to do more than we can actually do, and underestimate how long it takes to perform each task. Decision makers can benefit from aids that identify when these failure situations are likely, the root causes of these failures, and resolutions to these failures. In this paper, we present a decision assistant that helps users resolve over-subscribed temporal problems. The system works like an experienced advisor that can quickly identify the cause of failure underlying temporal problems and compute resolutions. The core of the decision assistant is the Best-first Conflict-Directed Relaxation (BCDR) algorithm, which can detect conflicting sets of constraints within temporal problems, and computes continuous relaxations for them that weaken constraints to the minimum extent, instead of removing them completely. BCDR is an extension to the Conflict-Directed A* algorithm, first developed in the model-based reasoning community to compute most likely system diagnoses or reconfigurations. It generalizes the discrete conflicts and relaxations, to hybrid conflicts and relaxations, which denote minimal inconsistencies and minimal relaxations to both discrete and continuous relaxable constraints. In addition, BCDR is capable of handling temporal uncertainty, expressed as either set-bounded or probabilistic durations, and can compute preferred trade-offs between the risk of violating a schedule requirement, versus the loss of utility by weakening those requirements. BCDR has been applied to several decision support applications in different domains, including deep-sea exploration, urban travel planning and transit system management. It has demonstrated its effectiveness in helping users resolve over-subscribed scheduling problems and evaluate the robustness of existing solutions. In our benchmark experiments, BCDR has also demonstrated its efficiency on solving large-scale scheduling problems in the aforementioned domains. Thanks to its conflict-driven approach for computing relaxations, BCDR achieves one to two orders of magnitude improvements on runtime performance when compared to state-of-the-art numerical solvers.

Headlines today are littered with references to'the fourth industrial revolution', 'artificial intelligence', 'machine learning' and'big data'. While this hype isn't unfounded, the practical ways of achieving value often remain unclear to the industry. The bottom line is a primary concern for many manufacturers. Not without reason; many associate the fourth industrial revolution with further expenditure - instalment of new sensors to collect data, investment in data storage or perhaps in 3D printing equipment – rather than cost savings. Fortunately, the reality doesn't always need not be this expensive.

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.

Robot-assisted dressing offers an opportunity to benefit the lives of many people with disabilities, such as some older adults. However, robots currently lack common sense about the physical implications of their actions on people. The physical implications of dressing are complicated by non-rigid garments, which can result in a robot indirectly applying high forces to a person's body. We present a deep recurrent model that, when given a proposed action by the robot, predicts the forces a garment will apply to a person's body. We also show that a robot can provide better dressing assistance by using this model with model predictive control. The predictions made by our model only use haptic and kinematic observations from the robot's end effector, which are readily attainable. Collecting training data from real world physical human-robot interaction can be time consuming, costly, and put people at risk. Instead, we train our predictive model using data collected in an entirely self-supervised fashion from a physics-based simulation. We evaluated our approach with a PR2 robot that attempted to pull a hospital gown onto the arms of 10 human participants. With a 0.2s prediction horizon, our controller succeeded at high rates and lowered applied force while navigating the garment around a persons fist and elbow without getting caught. Shorter prediction horizons resulted in significantly reduced performance with the sleeve catching on the participants' fists and elbows, demonstrating the value of our model's predictions. These behaviors of mitigating catches emerged from our deep predictive model and the controller objective function, which primarily penalizes high forces.

We propose a novel method that makes use of deep neural networks and gradient decent to perform automated design on complex real world engineering tasks. Our approach works by training a neural network to mimic the fitness function of a design optimization task and then, using the differential nature of the neural network, perform gradient decent to maximize the fitness. We demonstrate this methods effectiveness by designing an optimized heat sink and both 2D and 3D airfoils that maximize the lift drag ratio under steady state flow conditions. We highlight that our method has two distinct benefits over other automated design approaches. First, evaluating the neural networks prediction of fitness can be orders of magnitude faster then simulating the system of interest. Second, using gradient decent allows the design space to be searched much more efficiently then other gradient free methods. These two strengths work together to overcome some of the current shortcomings of automated design.

Ensuring stable, secure and reliable operations of the power grid is a primary concern for system operators [1]. Security assessment and control actions heavily rely on the accuracy of the assumed power system model and its parameters and of the estimated state [2]. Thus, inaccuracies in state estimation data or in the networked dynamic model can impact the assessment of the system stability and the efficacy of the corresponding control measures. In this paper, we explore the possibility to leverage the proliferation of Phasor Measurement Units (PMUs) that collect time synchronous data in a distributed way, for validating the assumed power system model and the current system state. In particular, our goal is to develop a data-efficient learning framework for performing an online reconstruction of the dynamic model using the minimal number of assumptions and exclusively relying on the PMU measurements. A number of recent works showed promising results in attacking this problem [3], [4], [5], [6], [7], [8], [9]. Here, we propose to extend the scope of existing works to the problem of extracting the dynamic state matrix from PMU measurements in a purely data-driven way, without assuming any knowledge of model parameters. We take advantage of the separation of scales that exists in the regime of ambient fluctuations around the steady state leading to power system dynamics excited by stochastic load variations.

CIOs looking for a digital transformation case study will find an ongoing master class at GE. You forgot to provide an Email Address. This email address doesn't appear to be valid. This email address is already registered. You have exceeded the maximum character limit.

Machine learning has returned with a vengeance. I still remember the dark days of the late '80s and '90s, when it was pretty clear that the current generation of machine-learning algorithms didn't seem to actually learn much of anything. Then big data arrived, computers became chess geniuses, conquered Go (twice), and started recommending sentences to judges. In most of these cases, the computer had sucked up vast reams of data and created models based on the correlations in the data. It seems that quantum machine learning might provide an advantage here, as a recent paper on searching for Higgs bosons in particle physics data seems to hint.