Recent projections by the US federal government estimate that there will be 200,000 new coronavirus cases in the US by June 1. At the same time, governments around the world are grappling with the complexities of safely reopening businesses, schools and other public institutions. Technology companies are rushing into that gap with software aimed at keeping people safe, while citizens navigate a patchwork approach to easing shelter-in-place orders. One well-known approach is the use of contact-tracing apps on smart phones created by tech and telecom companies. These apps alert people if they've been in close proximity to an infected person.

Bit by bit, byte by byte, artificial intelligence has been working its way into public consciousness and into everyday computer use. Artificial intelligence and deep learning have been deeply woven into more and more aspects of end-user computing. Smartphones and other mobile devices use AI as well. Up until now, the artificial intelligence work has been done in the cloud, but a new approach to software design aims to arm mobile devices with real artificial-intelligence capability. "A mobile device is very resource-constrained," explained William & Mary computer scientist Bin Ren.

Machines can learn a lot of things--probably more than you can imagine. But can they learn common sense? At his company, Elemental Cognition, Ferrucci described how his AI team gave an advanced language program the sentence, "Zoey moves her plant to a sunny window. The AI program was tasked to complete the second sentence. SEE: An IT pro's guide to robotic process automation (free PDF) (TechRepublic) A human would likely complete the sentence by saying, "the sun will help the plant to grow and stay healthy." In the real world, it's common knowledge that plants need light. Unfortunately, the AI program couldn't deliver this common observation. Instead, the AI completed the sentence by analyzing statistical patterns. It came up with these possible answers: "she finds something, not pleasant," "fertilizer is visible in the window," and "another plant is missing from the bedroom." This story is an entry point to myriad "common sense" issues that face today's AI. It begins to explain why a self-driving vehicle may not be able to decipher the varying degrees of danger between striking a traffic cone or striking a pedestrian. "The great irony of common sense--and indeed AI itself--is that it is stuff that pretty much everybody knows, yet nobody seems to know what exactly it is or how to build machines that possess it," said Gary Marcus, CEO and founder of Robust.AI. "Solving this problem is, we would argue, the single most important step towards taking AI to the next level.

UK GDP fell by 2 per cent in the first quarter of 2020, the most rapid contraction of the UK's economy since the 2008 financial crisis. Rishi Sunak, the chancellor of the exchequer, said, "It is now very likely that the UK economy will face a significant recession this year, and we're already in the middle of that as we speak." The Bank of England predicts that the UK economy could shrink by as much as 14 per cent in 2020. In England some people who aren't able to work from home returned to work today, as part of the government's recent easing of certain restrictions. Despite the government urging people to avoid public transport if they could, some commuters said buses and trains were too crowded to practice social distancing. It could be as long as "four or five years" before covid-19 is under control and the pandemic could "potentially get worse", according to the World Health Organization's chief scientist Soumya Swaminathan. Speaking at an FT conference, she said a vaccine "seems ...

Recommendation systems have lately been popularized globally, with primary use cases in online interaction systems, with significant focus on e-commerce platforms. We have developed a machine learning-based recommendation platform, which can be easily applied to almost any items and/or actions domain. Contrary to existing recommendation systems, our platform supports multiple types of interaction data with multiple modalities of metadata natively. This is achieved through multi-modal fusion of various data representations. We deployed the platform into multiple e-commerce stores of different kinds, e.g. food and beverages, shoes, fashion items, telecom operators. Here, we present our system, its flexibility and performance. We also show benchmark results on open datasets, that significantly outperform state-of-the-art prior work.

Machine learning systems have been shown to propagate the societal errors of the past. In light of this, a wealth of research focuses on designing solutions that are "fair." Even with this abundance of work, there is no singular definition of fairness, mainly because fairness is subjective and context dependent. We propose a new fairness definition, motivated by the principle of equity, that considers existing biases in the data and attempts to make equitable decisions that account for these previous historical biases. We formalize our definition of fairness, and motivate it with its appropriate contexts. Next, we operationalize it for equitable classification. We perform multiple automatic and human evaluations to show the effectiveness of our definition and demonstrate its utility for aspects of fairness, such as the feedback loop.

F-systems are digraphs that enable to model sentences that predicate the falsity of other sentences. Paradoxes like the Liar and Yablo's can be analyzed with that tool to find graph-theoretic patterns. In this paper we present the F-systems model abstracting from all the features of the language in which the represented sentences are expressed. All that is assumed is the existence of sentences and the binary relation '... affirms the falsity of ...' among them. The possible existence of non-referential sentences is also considered. To model the sets of all the sentences that can jointly be valued as true we introduce the notion of conglomerate, the existence of which guarantees the absence of paradox. Conglomerates also enable to characterize referential contradictions, i.e. sentences that can only be false under a classical valuation due to the interactions with other sentences in the model. A Kripke's style fixed point characterization of groundedness is offered and fixed points which are complete (meaning that every sentence is deemed either true or false) and consistent (meaning that no sentence is deemed true and false) are put in correspondence with conglomerates. Furthermore, argumentation frameworks are special cases of F-systems. We show the relation between local conglomerates and admissible sets of arguments and argue about the usefulness of the concept for argumentation theory.

We present a novel network pruning algorithm called Dynamic Sparse Training that can jointly find the optimal network parameters and sparse network structure in a unified optimization process with trainable pruning thresholds. These thresholds can have fine-grained layer-wise adjustments dynamically via backpropagation. We demonstrate that our dynamic sparse training algorithm can easily train very sparse neural network models with little performance loss using the same number of training epochs as dense models. Dynamic Sparse Training achieves state of the art performance compared with other sparse training algorithms on various network architectures. Additionally, we have several surprising observations that provide strong evidence to the effectiveness and efficiency of our algorithm. These observations reveal the underlying problems of traditional three-stage pruning algorithms and present the potential guidance provided by our algorithm to the design of more compact network architectures. Despite the impressive success that deep neural networks have achieved in a wide range of challenging tasks, the inference in deep neural networks is highly memory-intensive and computationintensive due to the over-parameterization of deep neural networks. Network pruning (LeCun et al. (1990); Han et al. (2015); Molchanov et al. (2017)) has been recognized as an effective approach to improving the inference efficiency in resource-limited scenarios. Traditional pruning methods consist of dense network training followed with pruning and fine-tuning iterations. To avoid the expensive pruning and fine-tuning iterations, many sparse training methods (Mocanu et al., 2018; Bellec et al., 2017; Mostafa & Wang, 2019; Dettmers & Zettlemoyer, 2019) have been proposed, where the network pruning is conducted during the training process. However, all these methods suffer from following three problems: Coarse-grained predefined pruning schedule.

Due to significant improvements in performance in recent years, neural networks are currently used for an ever-increasing number of applications. However, neural networks have the drawback that their decisions are not readily interpretable and traceable for a human. This creates several problems, for instance in terms of safety and IT security for high-risk applications, where assuring these properties is crucial. One of the most striking IT security problems aggravated by the opacity of neural networks is the possibility of so-called poisoning attacks during the training phase, where an attacker inserts specially crafted data to manipulate the resulting model. We propose an approach to this problem which allows provably verifying the integrity of the training procedure by making use of standard cryptographic mechanisms.

Efficiency criteria for conformal prediction, such as \emph{observed fuzziness} (i.e., the sum of p-values associated with false labels), are commonly used to \emph{evaluate} the performance of given conformal predictors. Here, we investigate whether it is possible to exploit efficiency criteria to \emph{learn} classifiers, both conformal predictors and point classifiers, by using such criteria as training objective functions. The proposed idea is implemented for the problem of binary classification of hand-written digits. By choosing a 1-dimensional model class (with one real-valued free parameter), we can solve the optimization problems through an (approximate) exhaustive search over (a discrete version of) the parameter space. Our empirical results suggest that conformal predictors trained by minimizing their observed fuzziness perform better than conformal predictors trained in the traditional way by minimizing the \emph{prediction error} of the corresponding point classifier. They also have a reasonable performance in terms of their prediction error on the test set.