The increasing impact of algorithmic decisions on people's lives compels us to scrutinize their fairness and, in particular, the disparate impacts that ostensibly-color-blind algorithms can have on different groups. Examples include credit decisioning, hiring, advertising, criminal justice, personalized medicine, and targeted policymaking, where in some cases legislative or regulatory frameworks for fairness exist and define specific protected classes. In this paper we study a fundamental challenge to assessing disparate impacts in practice: protected class membership is often not observed in the data. This is particularly a problem in lending and healthcare. We consider the use of an auxiliary dataset, such as the US census, that includes class labels but not decisions or outcomes. We show that a variety of common disparity measures are generally unidentifiable aside for some unrealistic cases, providing a new perspective on the documented biases of popular proxy-based methods. We provide exact characterizations of the sharpest-possible partial identification set of disparities either under no assumptions or when we incorporate mild smoothness constraints. We further provide optimization-based algorithms for computing and visualizing these sets, which enables reliable and robust assessments -- an important tool when disparity assessment can have far-reaching policy implications. We demonstrate this in two case studies with real data: mortgage lending and personalized medicine dosing.

Deep neural networks (DNNs) have recently achieved state-of-the-art performance and provide significant progress in many machine learning tasks, such as image classification, speech processing, natural language processing, etc. However, recent studies have shown that DNNs are vulnerable to adversarial attacks. For instance, in the image classification domain, adding small imperceptible perturbations to the input image is sufficient to fool the DNN and to cause misclassification. The perturbed image, called \textit{adversarial example}, should be visually as close as possible to the original image. However, all the works proposed in the literature for generating adversarial examples have used the $L_{p}$ norms ($L_{0}$, $L_{2}$ and $L_{\infty}$) as distance metrics to quantify the similarity between the original image and the adversarial example. Nonetheless, the $L_{p}$ norms do not correlate with human judgment, making them not suitable to reliably assess the perceptual similarity/fidelity of adversarial examples. In this paper, we present a database for visual fidelity assessment of adversarial examples. We describe the creation of the database and evaluate the performance of fifteen state-of-the-art full-reference (FR) image fidelity assessment metrics that could substitute $L_{p}$ norms. The database as well as subjective scores are publicly available to help designing new metrics for adversarial examples and to facilitate future research works.

Former Prime Minister David Cameron has been hired by a US artificial intelligence company. Mr Cameron, who quit UK politics in the wake of the 2016 EU Referendum, will lead Afiniti's advisory board. The firm said his "experience of political leadership through periods of complexity" would be invaluable. Mr Cameron has kept a relatively low profile since leaving UK politics - his autobiography is due to be published in September. He is also president of Alzheimer's Research UK and chairs the board of patrons at National Citizen Service.

BEIJING - Thirty years after the crackdown on Tiananmen protesters, the tanks that lined Beijing's central avenue have been replaced by countless surveillance cameras perched like hawks on lampposts to keep the population in check. The Chinese Communist Party has gone to great lengths to prevent another pro-democracy movement, clamping down on student activists, labor movements and lawyers with the help of high-tech surveillance. But the party has also pushed economic reforms that have made millions of people wealthier -- and less interested in rebelling like the students whose protest ended with hundreds killed on June 4, 1989. Over the past decade, small police booths have been set up block by block across the country to monitor neighborhood disputes, prevent crime, and keep tabs on anyone suspected of disturbing social order. Now China's obsession with artificial intelligence and facial recognition adds another layer of sophistication to this intricate surveillance web, allowing police to pound on the door of any perceived troublemaker, several activists have said.

Much has been made in recent years about the explosion of artificial intelligence (AI) in radiology and how it might impact the role of radiologists themselves. But artificial intelligence is, by definition, artificial. In an itnTV video from the 2018 Radiological Society of North America (RSNA) annual meeting, ITN Contributing Editor Greg Freiherr explored how AI cannot take the place of people, but it can help people get what they need. You can view the video at https://bit.ly/2FMgDvH. "I doubt any radiologist could build an MR or a CT scanner from scratch. They probably couldn't even build it from pieces," said Bradley J. Erickson, M.D., Ph.D., chair, radiology informatics/associate chair, research-radiology at the Mayo Clinic in Rochester, Minn., in the itnTV video.

On May 22, Congress held a hearing on facial recognition technology. Here are 4 things everyone needs to know about facial recognition tech. On May 22, the House Oversight and Reform Committee held a hearing on facial recognition technology. The Congressional hearing examines the use of facial recognition technology by the federal government, businesses, social media platforms and law enforcement. Specifically, lawmakers address privacy and civil rights concerns.

Anti-Virus software is the layman's solution to cybersecurity. Functioning as a first line of defense, Anti-Virus software works to prevent, detect, and remove malware from your computer. However, Anti-Virus software is not a cure all solution. In fact, IMB Knowledge Center published a piece on the limitations of Anti-Virus protection. In the article, they cite file size, scan time, and nesting depth as a few of the limitations.

Anti-Virus software is the layman's solution to cybersecurity. Functioning as a first line of defense, Anti-Virus software works to prevent, detect, and remove malware from your computer. However, Anti-Virus software is not a cure all solution. In fact, IMB Knowledge Center published a piece on the limitations of Anti-Virus protection. In the article, they cite file size, scan time, and nesting depth as a few of the limitations.

Spatial-temporal graph modeling is an important task to analyze the spatial relations and temporal trends of components in a system. Existing approaches mostly capture the spatial dependency on a fixed graph structure, assuming that the underlying relation between entities is pre-determined. However, the explicit graph structure (relation) does not necessarily reflect the true dependency and genuine relation may be missing due to the incomplete connections in the data. Furthermore, existing methods are ineffective to capture the temporal trends as the RNNs or CNNs employed in these methods cannot capture long-range temporal sequences. To overcome these limitations, we propose in this paper a novel graph neural network architecture, Graph WaveNet, for spatial-temporal graph modeling. By developing a novel adaptive dependency matrix and learn it through node embedding, our model can precisely capture the hidden spatial dependency in the data. With a stacked dilated 1D convolution component whose receptive field grows exponentially as the number of layers increases, Graph WaveNet is able to handle very long sequences. These two components are integrated seamlessly in a unified framework and the whole framework is learned in an end-to-end manner. Experimental results on two public traffic network datasets, METR-LA and PEMS-BAY, demonstrate the superior performance of our algorithm.

Recently there has been a significant interest in learning disentangled representations, as they promise increased interpretability, generalization to unseen scenarios and faster learning on downstream tasks. In this paper, we investigate the usefulness of different notions of disentanglement for improving the fairness of downstream prediction tasks based on representations. We consider the setting where the goal is to predict a target variable based on the learned representation of high-dimensional observations (such as images) that depend on both the target variable and an unobserved sensitive variable. We show that in this setting both the optimal and empirical predictions can be unfair, even if the target variable and the sensitive variable are independent. Analyzing more than 12 600 trained representations of state-of-the-art disentangled models, we observe that various disentanglement scores are consistently correlated with increased fairness, suggesting that disentanglement may be a useful property to encourage fairness when sensitive variables are not observed.