High-accurate machine learning (ML) image classifiers cannot guarantee that they will not fail at operation. Thus, their deployment in safety-critical applications such as autonomous vehicles is still an open issue. The use of fault tolerance mechanisms such as safety monitors is a promising direction to keep the system in a safe state despite errors of the ML classifier. As the prediction from the ML is the core information directly impacting safety, many works are focusing on monitoring the ML model itself. Checking the efficiency of such monitors in the context of safety-critical applications is thus a significant challenge. Therefore, this paper aims at establishing a baseline framework for benchmarking monitors for ML image classifiers. Furthermore, we propose a framework covering the entire pipeline, from data generation to evaluation. Our approach measures monitor performance with a broader set of metrics than usually proposed in the literature. Moreover, we benchmark three different monitor approaches in 79 benchmark datasets containing five categories of out-of-distribution data for image classifiers: class novelty, noise, anomalies, distributional shifts, and adversarial attacks. Our results indicate that these monitors are no more accurate than a random monitor. We also release the code of all experiments for reproducibility.

Fast developing artificial intelligence (AI) technology has enabled various applied systems deployed in the real world, impacting people's everyday lives. However, many current AI systems were found vulnerable to imperceptible attacks, biased against underrepresented groups, lacking in user privacy protection, etc., which not only degrades user experience but erodes the society's trust in all AI systems. In this review, we strive to provide AI practitioners a comprehensive guide towards building trustworthy AI systems. We first introduce the theoretical framework of important aspects of AI trustworthiness, including robustness, generalization, explainability, transparency, reproducibility, fairness, privacy preservation, alignment with human values, and accountability. We then survey leading approaches in these aspects in the industry. To unify the current fragmented approaches towards trustworthy AI, we propose a systematic approach that considers the entire lifecycle of AI systems, ranging from data acquisition to model development, to development and deployment, finally to continuous monitoring and governance. In this framework, we offer concrete action items to practitioners and societal stakeholders (e.g., researchers and regulators) to improve AI trustworthiness. Finally, we identify key opportunities and challenges in the future development of trustworthy AI systems, where we identify the need for paradigm shift towards comprehensive trustworthy AI systems.

In this paper, we study the learning problem in contextual search, which is motivated by applications such as first-price auction, personalized medicine experiments, and feature-based pricing experiments. In particular, for a sequence of arriving context vectors, with each context associated with an underlying value, the decision-maker either makes a query at a certain point or skips the context. The decision-maker will only observe the binary feedback on the relationship between the query point and the value associated with the context. We study a PAC learning setting, where the goal is to learn the underlying mean value function in context with a minimum number of queries. To address this challenge, we propose a tri-section search approach combined with a margin-based active learning method. We show that the algorithm only needs to make $O(1/\varepsilon^2)$ queries to achieve an $\epsilon$-estimation accuracy. This sample complexity significantly reduces the required sample complexity in the passive setting, at least $\Omega(1/\varepsilon^4)$.

Treeging combines the flexible mean structure of regression trees with the covariance-based prediction strategy of kriging into the base learner of an ensemble prediction algorithm. In so doing, it combines the strengths of the two primary types of spatial and space-time prediction models: (1) models with flexible mean structures (often machine learning algorithms) that assume independently distributed data, and (2) kriging or Gaussian Process (GP) prediction models with rich covariance structures but simple mean structures. We investigate the predictive accuracy of treeging across a thorough and widely varied battery of spatial and space-time simulation scenarios, comparing it to ordinary kriging, random forest and ensembles of ordinary kriging base learners. Treeging performs well across the board, whereas kriging suffers when dependence is weak or in the presence of spurious covariates, and random forest suffers when the covariates are less informative. Treeging also outperforms these competitors in predicting atmospheric pollutants (ozone and PM$_{2.5}$) in several case studies. We examine sensitivity to tuning parameters (number of base learners and training data sampling proportion), finding they follow the familiar intuition of their random forest counterparts. We include a discussion of scaleability, noting that any covariance approximation techniques that expedite kriging (GP) may be similarly applied to expedite treeging.

Addenbrooke's Hospital in Cambridge along with 20 other hospitals from across the world and healthcare technology leader, NVIDIA, have used artificial intelligence (AI) to predict Covid patients' oxygen needs on a global scale. The research was sparked by the pandemic and set out to build an AI tool to predict how much extra oxygen a Covid-19 patient may need in the first days of hospital care, using data from across four continents. The technique, known as federated learning, used an algorithm to analyze chest x-rays and electronic health data from hospital patients with Covid symptoms. To maintain strict patient confidentiality, the patient data was fully anonymized and an algorithm was sent to each hospital so no data was shared or left its location. Once the algorithm had'learned' from the data, the analysis was brought together to build an AI tool which could predict the oxygen needs of hospital Covid patients anywhere in the world.

Graph convolutional neural networks (GCNNs) are popular deep learning architectures that, upon replacing regular convolutions with graph filters (GFs), generalize CNNs to irregular domains. However, classical GFs are prone to numerical errors since they consist of high-order polynomials. This problem is aggravated when several filters are applied in cascade, limiting the practical depth of GCNNs. To tackle this issue, we present the neighborhood graph filters (NGFs), a family of GFs that replaces the powers of the graph shift operator with $k$-hop neighborhood adjacency matrices. NGFs help to alleviate the numerical issues of traditional GFs, allow for the design of deeper GCNNs, and enhance the robustness to errors in the topology of the graph. To illustrate the advantage over traditional GFs in practical applications, we use NGFs in the design of deep neighborhood GCNNs to solve graph signal denoising and node classification problems over both synthetic and real-world data.

RIO DE JANEIRO, BRAZIL - The Brazilian Chamber of Deputies on Wednesday, September 29, passed a bill establishing principles and guidelines for the development and use of artificial intelligence (the ability of a machine or system to learn commands on its own and use them). Using this kind of technology to violate the protection of free competition or to engage in abusive market practices will be forbidden. The text will now be debated by the Senate. Deputy Eduardo Bismarck's bill determines that the government issue regulations with guidelines to monitor the risk management of artificial intelligence systems and establish the . . . To read the full NEWS and much more, Subscribe to our Premium Membership Plan.

Mavenir, the Network Software Provider building the future of networks with cloud-native software that runs on any cloud and transforms the way the world connects, announced its next-generation Operations Support System (ngOSS), a new addition to the MAVscale platform. The new solution brings together a powerful combination of Artificial Intelligence (AI), Analytics, Automation, and Orchestration, backed by Mavenir's experience delivering telecom network solutions to allow Communication Service Providers (CSPs) to deploy ngOSS in their 4G networks today, while at the same time preparing to fully realize the potential of 5G capabilities through intelligent network operations. By enhancing network performance, Mavenir's ngOSS enables CSPs to improve the overall customer experience, lower operating expenses (OPEX), and reduce the risk of manual errors more prevalent in legacy OSS systems. The end-to-end, cloud-native automation framework is built based on TM Forum's Open Digital Architecture (ODA) and OpenAPIs that uses enhanced AI and Machine Learning (ML) models to deliver advanced network performance. A more open, automated network is necessary to break legacy vendor lock-in while driving innovation for 5G use cases.

Verdict lists the top five terms tweeted on future cities in Q2 2021, based on data from GlobalData's Influencer Platform. The top tweeted terms are the trending industry discussions happening on Twitter by key individuals (influencers) as tracked by the platform. The importance of analytics and visualisations for city governments, how smart cities can unlock the full potential of IoT, and the applications of IoT in smart cities and smart buildings were some of the trending discussions in Q2. Kirk Borne, data scientist at DataPrime, an artificial intelligence (AI)-based solutions provider, shared an article on the importance of analytics and visualisation for city governments. Government policymakers are required to collect and store data to ensure data privacy across organisations and to enable the provision of services that boost the economy.

Much of the world's population experiences some form of disability during their lifetime. Caution must be exercised while designing natural language processing (NLP) systems to prevent systems from inadvertently perpetuating ableist bias against people with disabilities, i.e., prejudice that favors those with typical abilities. We report on various analyses based on word predictions of a large-scale BERT language model. Statistically significant results demonstrate that people with disabilities can be disadvantaged. Findings also explore overlapping forms of discrimination related to interconnected gender and race identities.